diff options
Diffstat (limited to 'fs')
75 files changed, 1612 insertions, 864 deletions
diff --git a/fs/9p/vfs_dir.c b/fs/9p/vfs_dir.c index 16c8a2a98c1..899f168fd19 100644 --- a/fs/9p/vfs_dir.c +++ b/fs/9p/vfs_dir.c @@ -292,9 +292,11 @@ int v9fs_dir_release(struct inode *inode, struct file *filp) fid = filp->private_data; P9_DPRINTK(P9_DEBUG_VFS, - "inode: %p filp: %p fid: %d\n", inode, filp, fid->fid); + "v9fs_dir_release: inode: %p filp: %p fid: %d\n", + inode, filp, fid ? fid->fid : -1); filemap_write_and_wait(inode->i_mapping); - p9_client_clunk(fid); + if (fid) + p9_client_clunk(fid); return 0; } diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index c7c23eab944..9e670d52764 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -730,7 +730,10 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int mode, P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n", err); goto error; } - dentry->d_op = &v9fs_cached_dentry_operations; + if (v9ses->cache) + dentry->d_op = &v9fs_cached_dentry_operations; + else + dentry->d_op = &v9fs_dentry_operations; d_instantiate(dentry, inode); err = v9fs_fid_add(dentry, fid); if (err < 0) @@ -1128,6 +1131,7 @@ v9fs_vfs_getattr(struct vfsmount *mnt, struct dentry *dentry, v9fs_stat2inode(st, dentry->d_inode, dentry->d_inode->i_sb); generic_fillattr(dentry->d_inode, stat); + p9stat_free(st); kfree(st); return 0; } @@ -1489,6 +1493,7 @@ static int v9fs_readlink(struct dentry *dentry, char *buffer, int buflen) retval = strnlen(buffer, buflen); done: + p9stat_free(st); kfree(st); return retval; } @@ -1942,7 +1947,7 @@ static const struct inode_operations v9fs_dir_inode_operations_dotu = { .unlink = v9fs_vfs_unlink, .mkdir = v9fs_vfs_mkdir, .rmdir = v9fs_vfs_rmdir, - .mknod = v9fs_vfs_mknod_dotl, + .mknod = v9fs_vfs_mknod, .rename = v9fs_vfs_rename, .getattr = v9fs_vfs_getattr, .setattr = v9fs_vfs_setattr, diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index f9311077de6..1d12ba0ed3d 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -122,6 +122,10 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, fid = v9fs_session_init(v9ses, dev_name, data); if (IS_ERR(fid)) { retval = PTR_ERR(fid); + /* + * we need to call session_close to tear down some + * of the data structure setup by session_init + */ goto close_session; } @@ -144,7 +148,6 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, retval = -ENOMEM; goto release_sb; } - sb->s_root = root; if (v9fs_proto_dotl(v9ses)) { @@ -152,7 +155,7 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, st = p9_client_getattr_dotl(fid, P9_STATS_BASIC); if (IS_ERR(st)) { retval = PTR_ERR(st); - goto clunk_fid; + goto release_sb; } v9fs_stat2inode_dotl(st, root->d_inode); @@ -162,7 +165,7 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, st = p9_client_stat(fid); if (IS_ERR(st)) { retval = PTR_ERR(st); - goto clunk_fid; + goto release_sb; } root->d_inode->i_ino = v9fs_qid2ino(&st->qid); @@ -174,19 +177,24 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, v9fs_fid_add(root, fid); -P9_DPRINTK(P9_DEBUG_VFS, " simple set mount, return 0\n"); + P9_DPRINTK(P9_DEBUG_VFS, " simple set mount, return 0\n"); simple_set_mnt(mnt, sb); return 0; clunk_fid: p9_client_clunk(fid); - close_session: v9fs_session_close(v9ses); kfree(v9ses); return retval; - release_sb: + /* + * we will do the session_close and root dentry release + * in the below call. But we need to clunk fid, because we haven't + * attached the fid to dentry so it won't get clunked + * automatically. + */ + p9_client_clunk(fid); deactivate_locked_super(sb); return retval; } @@ -712,8 +712,16 @@ static ssize_t aio_run_iocb(struct kiocb *iocb) */ ret = retry(iocb); - if (ret != -EIOCBRETRY && ret != -EIOCBQUEUED) + if (ret != -EIOCBRETRY && ret != -EIOCBQUEUED) { + /* + * There's no easy way to restart the syscall since other AIO's + * may be already running. Just fail this IO with EINTR. + */ + if (unlikely(ret == -ERESTARTSYS || ret == -ERESTARTNOINTR || + ret == -ERESTARTNOHAND || ret == -ERESTART_RESTARTBLOCK)) + ret = -EINTR; aio_complete(iocb, ret, 0); + } out: spin_lock_irq(&ctx->ctx_lock); @@ -1659,6 +1667,9 @@ long do_io_submit(aio_context_t ctx_id, long nr, if (unlikely(nr < 0)) return -EINVAL; + if (unlikely(nr > LONG_MAX/sizeof(*iocbpp))) + nr = LONG_MAX/sizeof(*iocbpp); + if (unlikely(!access_ok(VERIFY_READ, iocbpp, (nr*sizeof(*iocbpp))))) return -EFAULT; diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c index a7528b91393..fd0cc0bf9a4 100644 --- a/fs/binfmt_misc.c +++ b/fs/binfmt_misc.c @@ -724,7 +724,7 @@ static int __init init_misc_binfmt(void) { int err = register_filesystem(&bm_fs_type); if (!err) { - err = register_binfmt(&misc_format); + err = insert_binfmt(&misc_format); if (err) unregister_filesystem(&bm_fs_type); } diff --git a/fs/bio-integrity.c b/fs/bio-integrity.c index 612a5c38d3c..4d0ff5ee27b 100644 --- a/fs/bio-integrity.c +++ b/fs/bio-integrity.c @@ -413,10 +413,10 @@ int bio_integrity_prep(struct bio *bio) /* Allocate kernel buffer for protection data */ len = sectors * blk_integrity_tuple_size(bi); - buf = kmalloc(len, GFP_NOIO | __GFP_NOFAIL | q->bounce_gfp); + buf = kmalloc(len, GFP_NOIO | q->bounce_gfp); if (unlikely(buf == NULL)) { printk(KERN_ERR "could not allocate integrity buffer\n"); - return -EIO; + return -ENOMEM; } end = (((unsigned long) buf) + len + PAGE_SIZE - 1) >> PAGE_SHIFT; diff --git a/fs/ceph/Kconfig b/fs/ceph/Kconfig index bc87b9c1d27..0fcd2640c23 100644 --- a/fs/ceph/Kconfig +++ b/fs/ceph/Kconfig @@ -3,6 +3,7 @@ config CEPH_FS depends on INET && EXPERIMENTAL select LIBCRC32C select CRYPTO_AES + select CRYPTO help Choose Y or M here to include support for mounting the experimental Ceph distributed file system. Ceph is an extremely diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index 4cfce1ee31f..efbc604001c 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -411,8 +411,8 @@ static int writepage_nounlock(struct page *page, struct writeback_control *wbc) if (i_size < page_off + len) len = i_size - page_off; - dout("writepage %p page %p index %lu on %llu~%u\n", - inode, page, page->index, page_off, len); + dout("writepage %p page %p index %lu on %llu~%u snapc %p\n", + inode, page, page->index, page_off, len, snapc); writeback_stat = atomic_long_inc_return(&client->writeback_count); if (writeback_stat > @@ -766,7 +766,8 @@ get_more_pages: /* ok */ if (locked_pages == 0) { /* prepare async write request */ - offset = page->index << PAGE_CACHE_SHIFT; + offset = (unsigned long long)page->index + << PAGE_CACHE_SHIFT; len = wsize; req = ceph_osdc_new_request(&client->osdc, &ci->i_layout, diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index a2069b6680a..73c153092f7 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -814,7 +814,7 @@ int __ceph_caps_used(struct ceph_inode_info *ci) used |= CEPH_CAP_PIN; if (ci->i_rd_ref) used |= CEPH_CAP_FILE_RD; - if (ci->i_rdcache_ref || ci->i_rdcache_gen) + if (ci->i_rdcache_ref || ci->vfs_inode.i_data.nrpages) used |= CEPH_CAP_FILE_CACHE; if (ci->i_wr_ref) used |= CEPH_CAP_FILE_WR; @@ -1195,10 +1195,14 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap, * asynchronously back to the MDS once sync writes complete and dirty * data is written out. * + * Unless @again is true, skip cap_snaps that were already sent to + * the MDS (i.e., during this session). + * * Called under i_lock. Takes s_mutex as needed. */ void __ceph_flush_snaps(struct ceph_inode_info *ci, - struct ceph_mds_session **psession) + struct ceph_mds_session **psession, + int again) __releases(ci->vfs_inode->i_lock) __acquires(ci->vfs_inode->i_lock) { @@ -1227,7 +1231,7 @@ retry: * pages to be written out. */ if (capsnap->dirty_pages || capsnap->writing) - continue; + break; /* * if cap writeback already occurred, we should have dropped @@ -1240,6 +1244,13 @@ retry: dout("no auth cap (migrating?), doing nothing\n"); goto out; } + + /* only flush each capsnap once */ + if (!again && !list_empty(&capsnap->flushing_item)) { + dout("already flushed %p, skipping\n", capsnap); + continue; + } + mds = ci->i_auth_cap->session->s_mds; mseq = ci->i_auth_cap->mseq; @@ -1276,8 +1287,8 @@ retry: &session->s_cap_snaps_flushing); spin_unlock(&inode->i_lock); - dout("flush_snaps %p cap_snap %p follows %lld size %llu\n", - inode, capsnap, next_follows, capsnap->size); + dout("flush_snaps %p cap_snap %p follows %lld tid %llu\n", + inode, capsnap, capsnap->follows, capsnap->flush_tid); send_cap_msg(session, ceph_vino(inode).ino, 0, CEPH_CAP_OP_FLUSHSNAP, capsnap->issued, 0, capsnap->dirty, 0, capsnap->flush_tid, 0, mseq, @@ -1314,7 +1325,7 @@ static void ceph_flush_snaps(struct ceph_inode_info *ci) struct inode *inode = &ci->vfs_inode; spin_lock(&inode->i_lock); - __ceph_flush_snaps(ci, NULL); + __ceph_flush_snaps(ci, NULL, 0); spin_unlock(&inode->i_lock); } @@ -1477,7 +1488,7 @@ void ceph_check_caps(struct ceph_inode_info *ci, int flags, /* flush snaps first time around only */ if (!list_empty(&ci->i_cap_snaps)) - __ceph_flush_snaps(ci, &session); + __ceph_flush_snaps(ci, &session, 0); goto retry_locked; retry: spin_lock(&inode->i_lock); @@ -1894,7 +1905,7 @@ static void kick_flushing_capsnaps(struct ceph_mds_client *mdsc, if (cap && cap->session == session) { dout("kick_flushing_caps %p cap %p capsnap %p\n", inode, cap, capsnap); - __ceph_flush_snaps(ci, &session); + __ceph_flush_snaps(ci, &session, 1); } else { pr_err("%p auth cap %p not mds%d ???\n", inode, cap, session->s_mds); diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c index 6e4f43ff23e..a1986eb5204 100644 --- a/fs/ceph/dir.c +++ b/fs/ceph/dir.c @@ -1021,11 +1021,15 @@ out_touch: static void ceph_dentry_release(struct dentry *dentry) { struct ceph_dentry_info *di = ceph_dentry(dentry); - struct inode *parent_inode = dentry->d_parent->d_inode; - u64 snapid = ceph_snap(parent_inode); + struct inode *parent_inode = NULL; + u64 snapid = CEPH_NOSNAP; + if (!IS_ROOT(dentry)) { + parent_inode = dentry->d_parent->d_inode; + if (parent_inode) + snapid = ceph_snap(parent_inode); + } dout("dentry_release %p parent %p\n", dentry, parent_inode); - if (parent_inode && snapid != CEPH_SNAPDIR) { struct ceph_inode_info *ci = ceph_inode(parent_inode); diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index e7cca414da0..62377ec37ed 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -845,7 +845,7 @@ static void ceph_set_dentry_offset(struct dentry *dn) * the caller) if we fail. */ static struct dentry *splice_dentry(struct dentry *dn, struct inode *in, - bool *prehash) + bool *prehash, bool set_offset) { struct dentry *realdn; @@ -877,7 +877,8 @@ static struct dentry *splice_dentry(struct dentry *dn, struct inode *in, } if ((!prehash || *prehash) && d_unhashed(dn)) d_rehash(dn); - ceph_set_dentry_offset(dn); + if (set_offset) + ceph_set_dentry_offset(dn); out: return dn; } @@ -1062,7 +1063,7 @@ int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req, d_delete(dn); goto done; } - dn = splice_dentry(dn, in, &have_lease); + dn = splice_dentry(dn, in, &have_lease, true); if (IS_ERR(dn)) { err = PTR_ERR(dn); goto done; @@ -1105,7 +1106,7 @@ int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req, goto done; } dout(" linking snapped dir %p to dn %p\n", in, dn); - dn = splice_dentry(dn, in, NULL); + dn = splice_dentry(dn, in, NULL, true); if (IS_ERR(dn)) { err = PTR_ERR(dn); goto done; @@ -1237,7 +1238,7 @@ retry_lookup: err = PTR_ERR(in); goto out; } - dn = splice_dentry(dn, in, NULL); + dn = splice_dentry(dn, in, NULL, false); if (IS_ERR(dn)) dn = NULL; } diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index f091b135178..fad95f8f260 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -2374,6 +2374,8 @@ static int encode_caps_cb(struct inode *inode, struct ceph_cap *cap, num_fcntl_locks, num_flock_locks); unlock_kernel(); + } else { + err = ceph_pagelist_append(pagelist, &rec, reclen); } out_free: diff --git a/fs/ceph/pagelist.c b/fs/ceph/pagelist.c index b6859f47d36..46a368b6dce 100644 --- a/fs/ceph/pagelist.c +++ b/fs/ceph/pagelist.c @@ -5,10 +5,18 @@ #include "pagelist.h" +static void ceph_pagelist_unmap_tail(struct ceph_pagelist *pl) +{ + struct page *page = list_entry(pl->head.prev, struct page, + lru); + kunmap(page); +} + int ceph_pagelist_release(struct ceph_pagelist *pl) { if (pl->mapped_tail) - kunmap(pl->mapped_tail); + ceph_pagelist_unmap_tail(pl); + while (!list_empty(&pl->head)) { struct page *page = list_first_entry(&pl->head, struct page, lru); @@ -26,7 +34,7 @@ static int ceph_pagelist_addpage(struct ceph_pagelist *pl) pl->room += PAGE_SIZE; list_add_tail(&page->lru, &pl->head); if (pl->mapped_tail) - kunmap(pl->mapped_tail); + ceph_pagelist_unmap_tail(pl); pl->mapped_tail = kmap(page); return 0; } diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c index 4868b9dcac5..190b6c4a6f2 100644 --- a/fs/ceph/snap.c +++ b/fs/ceph/snap.c @@ -119,6 +119,7 @@ static struct ceph_snap_realm *ceph_create_snap_realm( INIT_LIST_HEAD(&realm->children); INIT_LIST_HEAD(&realm->child_item); INIT_LIST_HEAD(&realm->empty_item); + INIT_LIST_HEAD(&realm->dirty_item); INIT_LIST_HEAD(&realm->inodes_with_caps); spin_lock_init(&realm->inodes_with_caps_lock); __insert_snap_realm(&mdsc->snap_realms, realm); @@ -467,7 +468,7 @@ void ceph_queue_cap_snap(struct ceph_inode_info *ci) INIT_LIST_HEAD(&capsnap->ci_item); INIT_LIST_HEAD(&capsnap->flushing_item); - capsnap->follows = snapc->seq - 1; + capsnap->follows = snapc->seq; capsnap->issued = __ceph_caps_issued(ci, NULL); capsnap->dirty = dirty; @@ -604,6 +605,7 @@ int ceph_update_snap_trace(struct ceph_mds_client *mdsc, struct ceph_snap_realm *realm; int invalidate = 0; int err = -ENOMEM; + LIST_HEAD(dirty_realms); dout("update_snap_trace deletion=%d\n", deletion); more: @@ -626,24 +628,6 @@ more: } } - if (le64_to_cpu(ri->seq) > realm->seq) { - dout("update_snap_trace updating %llx %p %lld -> %lld\n", - realm->ino, realm, realm->seq, le64_to_cpu(ri->seq)); - /* - * if the realm seq has changed, queue a cap_snap for every - * inode with open caps. we do this _before_ we update - * the realm info so that we prepare for writeback under the - * _previous_ snap context. - * - * ...unless it's a snap deletion! - */ - if (!deletion) - queue_realm_cap_snaps(realm); - } else { - dout("update_snap_trace %llx %p seq %lld unchanged\n", - realm->ino, realm, realm->seq); - } - /* ensure the parent is correct */ err = adjust_snap_realm_parent(mdsc, realm, le64_to_cpu(ri->parent)); if (err < 0) @@ -651,6 +635,8 @@ more: invalidate += err; if (le64_to_cpu(ri->seq) > realm->seq) { + dout("update_snap_trace updating %llx %p %lld -> %lld\n", + realm->ino, realm, realm->seq, le64_to_cpu(ri->seq)); /* update realm parameters, snap lists */ realm->seq = le64_to_cpu(ri->seq); realm->created = le64_to_cpu(ri->created); @@ -668,9 +654,17 @@ more: if (err < 0) goto fail; + /* queue realm for cap_snap creation */ + list_add(&realm->dirty_item, &dirty_realms); + invalidate = 1; } else if (!realm->cached_context) { + dout("update_snap_trace %llx %p seq %lld new\n", + realm->ino, realm, realm->seq); invalidate = 1; + } else { + dout("update_snap_trace %llx %p seq %lld unchanged\n", + realm->ino, realm, realm->seq); } dout("done with %llx %p, invalidated=%d, %p %p\n", realm->ino, @@ -683,6 +677,14 @@ more: if (invalidate) rebuild_snap_realms(realm); + /* + * queue cap snaps _after_ we've built the new snap contexts, + * so that i_head_snapc can be set appropriately. + */ + list_for_each_entry(realm, &dirty_realms, dirty_item) { + queue_realm_cap_snaps(realm); + } + __cleanup_empty_realms(mdsc); return 0; @@ -715,7 +717,7 @@ static void flush_snaps(struct ceph_mds_client *mdsc) igrab(inode); spin_unlock(&mdsc->snap_flush_lock); spin_lock(&inode->i_lock); - __ceph_flush_snaps(ci, &session); + __ceph_flush_snaps(ci, &session, 0); spin_unlock(&inode->i_lock); iput(inode); spin_lock(&mdsc->snap_flush_lock); @@ -816,6 +818,7 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc, }; struct inode *inode = ceph_find_inode(sb, vino); struct ceph_inode_info *ci; + struct ceph_snap_realm *oldrealm; if (!inode) continue; @@ -841,18 +844,19 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc, dout(" will move %p to split realm %llx %p\n", inode, realm->ino, realm); /* - * Remove the inode from the realm's inode - * list, but don't add it to the new realm - * yet. We don't want the cap_snap to be - * queued (again) by ceph_update_snap_trace() - * below. Queue it _now_, under the old context. + * Move the inode to the new realm */ spin_lock(&realm->inodes_with_caps_lock); list_del_init(&ci->i_snap_realm_item); + list_add(&ci->i_snap_realm_item, + &realm->inodes_with_caps); + oldrealm = ci->i_snap_realm; + ci->i_snap_realm = realm; spin_unlock(&realm->inodes_with_caps_lock); spin_unlock(&inode->i_lock); - ceph_queue_cap_snap(ci); + ceph_get_snap_realm(mdsc, realm); + ceph_put_snap_realm(mdsc, oldrealm); iput(inode); continue; @@ -880,43 +884,9 @@ skip_inode: ceph_update_snap_trace(mdsc, p, e, op == CEPH_SNAP_OP_DESTROY); - if (op == CEPH_SNAP_OP_SPLIT) { - /* - * ok, _now_ add the inodes into the new realm. - */ - for (i = 0; i < num_split_inos; i++) { - struct ceph_vino vino = { - .ino = le64_to_cpu(split_inos[i]), - .snap = CEPH_NOSNAP, - }; - struct inode *inode = ceph_find_inode(sb, vino); - struct ceph_inode_info *ci; - - if (!inode) - continue; - ci = ceph_inode(inode); - spin_lock(&inode->i_lock); - if (list_empty(&ci->i_snap_realm_item)) { - struct ceph_snap_realm *oldrealm = - ci->i_snap_realm; - - dout(" moving %p to split realm %llx %p\n", - inode, realm->ino, realm); - spin_lock(&realm->inodes_with_caps_lock); - list_add(&ci->i_snap_realm_item, - &realm->inodes_with_caps); - ci->i_snap_realm = realm; - spin_unlock(&realm->inodes_with_caps_lock); - ceph_get_snap_realm(mdsc, realm); - ceph_put_snap_realm(mdsc, oldrealm); - } - spin_unlock(&inode->i_lock); - iput(inode); - } - + if (op == CEPH_SNAP_OP_SPLIT) /* we took a reference when we created the realm, above */ ceph_put_snap_realm(mdsc, realm); - } __cleanup_empty_realms(mdsc); diff --git a/fs/ceph/super.h b/fs/ceph/super.h index c33897ae572..b87638e84c4 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -690,6 +690,8 @@ struct ceph_snap_realm { struct list_head empty_item; /* if i have ref==0 */ + struct list_head dirty_item; /* if realm needs new context */ + /* the current set of snaps for this realm */ struct ceph_snap_context *cached_context; @@ -826,7 +828,8 @@ extern void ceph_put_cap_refs(struct ceph_inode_info *ci, int had); extern void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr, struct ceph_snap_context *snapc); extern void __ceph_flush_snaps(struct ceph_inode_info *ci, - struct ceph_mds_session **psession); + struct ceph_mds_session **psession, + int again); extern void ceph_check_caps(struct ceph_inode_info *ci, int flags, struct ceph_mds_session *session); extern void ceph_check_delayed_caps(struct ceph_mds_client *mdsc); diff --git a/fs/char_dev.c b/fs/char_dev.c index f80a4f25123..143d393881c 100644 --- a/fs/char_dev.c +++ b/fs/char_dev.c @@ -40,7 +40,9 @@ struct backing_dev_info directly_mappable_cdev_bdi = { #endif /* permit direct mmap, for read, write or exec */ BDI_CAP_MAP_DIRECT | - BDI_CAP_READ_MAP | BDI_CAP_WRITE_MAP | BDI_CAP_EXEC_MAP), + BDI_CAP_READ_MAP | BDI_CAP_WRITE_MAP | BDI_CAP_EXEC_MAP | + /* no writeback happens */ + BDI_CAP_NO_ACCT_AND_WRITEBACK), }; static struct kobj_map *cdev_map; diff --git a/fs/cifs/Kconfig b/fs/cifs/Kconfig index 0da1debd499..917b7d449bb 100644 --- a/fs/cifs/Kconfig +++ b/fs/cifs/Kconfig @@ -2,8 +2,6 @@ config CIFS tristate "CIFS support (advanced network filesystem, SMBFS successor)" depends on INET select NLS - select CRYPTO_MD5 - select CRYPTO_ARC4 help This is the client VFS module for the Common Internet File System (CIFS) protocol which is the successor to the Server Message Block diff --git a/fs/cifs/asn1.c b/fs/cifs/asn1.c index 21f0fbd8698..cfd1ce34e0b 100644 --- a/fs/cifs/asn1.c +++ b/fs/cifs/asn1.c @@ -597,13 +597,13 @@ decode_negTokenInit(unsigned char *security_blob, int length, if (compare_oid(oid, oidlen, MSKRB5_OID, MSKRB5_OID_LEN)) server->sec_mskerberos = true; - if (compare_oid(oid, oidlen, KRB5U2U_OID, + else if (compare_oid(oid, oidlen, KRB5U2U_OID, KRB5U2U_OID_LEN)) server->sec_kerberosu2u = true; - if (compare_oid(oid, oidlen, KRB5_OID, + else if (compare_oid(oid, oidlen, KRB5_OID, KRB5_OID_LEN)) server->sec_kerberos = true; - if (compare_oid(oid, oidlen, NTLMSSP_OID, + else if (compare_oid(oid, oidlen, NTLMSSP_OID, NTLMSSP_OID_LEN)) server->sec_ntlmssp = true; diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index 709f2296bdb..35042d8f733 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -27,7 +27,6 @@ #include "md5.h" #include "cifs_unicode.h" #include "cifsproto.h" -#include "ntlmssp.h" #include <linux/ctype.h> #include <linux/random.h> @@ -43,43 +42,21 @@ extern void SMBencrypt(unsigned char *passwd, const unsigned char *c8, unsigned char *p24); static int cifs_calculate_signature(const struct smb_hdr *cifs_pdu, - struct TCP_Server_Info *server, char *signature) + const struct mac_key *key, char *signature) { - int rc; + struct MD5Context context; - if (cifs_pdu == NULL || server == NULL || signature == NULL) + if ((cifs_pdu == NULL) || (signature == NULL) || (key == NULL)) return -EINVAL; - if (!server->ntlmssp.sdescmd5) { - cERROR(1, - "cifs_calculate_signature: can't generate signature\n"); - return -1; - } - - rc = crypto_shash_init(&server->ntlmssp.sdescmd5->shash); - if (rc) { - cERROR(1, "cifs_calculate_signature: oould not init md5\n"); - return rc; - } - - if (server->secType == RawNTLMSSP) - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, - server->session_key.data.ntlmv2.key, - CIFS_NTLMV2_SESSKEY_SIZE); - else - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, - (char *)&server->session_key.data, - server->session_key.len); - - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, - cifs_pdu->Protocol, cifs_pdu->smb_buf_length); + cifs_MD5_init(&context); + cifs_MD5_update(&context, (char *)&key->data, key->len); + cifs_MD5_update(&context, cifs_pdu->Protocol, cifs_pdu->smb_buf_length); - rc = crypto_shash_final(&server->ntlmssp.sdescmd5->shash, signature); - - return rc; + cifs_MD5_final(signature, &context); + return 0; } - int cifs_sign_smb(struct smb_hdr *cifs_pdu, struct TCP_Server_Info *server, __u32 *pexpected_response_sequence_number) { @@ -101,7 +78,8 @@ int cifs_sign_smb(struct smb_hdr *cifs_pdu, struct TCP_Server_Info *server, server->sequence_number++; spin_unlock(&GlobalMid_Lock); - rc = cifs_calculate_signature(cifs_pdu, server, smb_signature); + rc = cifs_calculate_signature(cifs_pdu, &server->mac_signing_key, + smb_signature); if (rc) memset(cifs_pdu->Signature.SecuritySignature, 0, 8); else @@ -111,39 +89,21 @@ int cifs_sign_smb(struct smb_hdr *cifs_pdu, struct TCP_Server_Info *server, } static int cifs_calc_signature2(const struct kvec *iov, int n_vec, - struct TCP_Server_Info *server, char *signature) + const struct mac_key *key, char *signature) { + struct MD5Context context; int i; - int rc; - if (iov == NULL || server == NULL || signature == NULL) + if ((iov == NULL) || (signature == NULL) || (key == NULL)) return -EINVAL; - if (!server->ntlmssp.sdescmd5) { - cERROR(1, "cifs_calc_signature2: can't generate signature\n"); - return -1; - } - - rc = crypto_shash_init(&server->ntlmssp.sdescmd5->shash); - if (rc) { - cERROR(1, "cifs_calc_signature2: oould not init md5\n"); - return rc; - } - - if (server->secType == RawNTLMSSP) - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, - server->session_key.data.ntlmv2.key, - CIFS_NTLMV2_SESSKEY_SIZE); - else - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, - (char *)&server->session_key.data, - server->session_key.len); - + cifs_MD5_init(&context); + cifs_MD5_update(&context, (char *)&key->data, key->len); for (i = 0; i < n_vec; i++) { if (iov[i].iov_len == 0) continue; if (iov[i].iov_base == NULL) { - cERROR(1, "cifs_calc_signature2: null iovec entry"); + cERROR(1, "null iovec entry"); return -EIO; } /* The first entry includes a length field (which does not get @@ -151,18 +111,18 @@ static int cifs_calc_signature2(const struct kvec *iov, int n_vec, if (i == 0) { if (iov[0].iov_len <= 8) /* cmd field at offset 9 */ break; /* nothing to sign or corrupt header */ - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, - iov[i].iov_base + 4, iov[i].iov_len - 4); + cifs_MD5_update(&context, iov[0].iov_base+4, + iov[0].iov_len-4); } else - crypto_shash_update(&server->ntlmssp.sdescmd5->shash, - iov[i].iov_base, iov[i].iov_len); + cifs_MD5_update(&context, iov[i].iov_base, iov[i].iov_len); } - rc = crypto_shash_final(&server->ntlmssp.sdescmd5->shash, signature); + cifs_MD5_final(signature, &context); - return rc; + return 0; } + int cifs_sign_smb2(struct kvec *iov, int n_vec, struct TCP_Server_Info *server, __u32 *pexpected_response_sequence_number) { @@ -185,7 +145,8 @@ int cifs_sign_smb2(struct kvec *iov, int n_vec, struct TCP_Server_Info *server, server->sequence_number++; spin_unlock(&GlobalMid_Lock); - rc = cifs_calc_signature2(iov, n_vec, server, smb_signature); + rc = cifs_calc_signature2(iov, n_vec, &server->mac_signing_key, + smb_signature); if (rc) memset(cifs_pdu->Signature.SecuritySignature, 0, 8); else @@ -195,14 +156,14 @@ int cifs_sign_smb2(struct kvec *iov, int n_vec, struct TCP_Server_Info *server, } int cifs_verify_signature(struct smb_hdr *cifs_pdu, - struct TCP_Server_Info *server, + const struct mac_key *mac_key, __u32 expected_sequence_number) { - int rc; + unsigned int rc; char server_response_sig[8]; char what_we_think_sig_should_be[20]; - if (cifs_pdu == NULL || server == NULL) + if ((cifs_pdu == NULL) || (mac_key == NULL)) return -EINVAL; if (cifs_pdu->Command == SMB_COM_NEGOTIATE) @@ -231,7 +192,7 @@ int cifs_verify_signature(struct smb_hdr *cifs_pdu, cpu_to_le32(expected_sequence_number); cifs_pdu->Signature.Sequence.Reserved = 0; - rc = cifs_calculate_signature(cifs_pdu, server, + rc = cifs_calculate_signature(cifs_pdu, mac_key, what_we_think_sig_should_be); if (rc) @@ -248,7 +209,7 @@ int cifs_verify_signature(struct smb_hdr *cifs_pdu, } /* We fill in key by putting in 40 byte array which was allocated by caller */ -int cifs_calculate_session_key(struct session_key *key, const char *rn, +int cifs_calculate_mac_key(struct mac_key *key, const char *rn, const char *password) { char temp_key[16]; @@ -306,52 +267,38 @@ static int calc_ntlmv2_hash(struct cifsSesInfo *ses, { int rc = 0; int len; - char nt_hash[CIFS_NTHASH_SIZE]; + char nt_hash[16]; + struct HMACMD5Context *pctxt; wchar_t *user; wchar_t *domain; - wchar_t *server; - if (!ses->server->ntlmssp.sdeschmacmd5) { - cERROR(1, "calc_ntlmv2_hash: can't generate ntlmv2 hash\n"); - return -1; - } + pctxt = kmalloc(sizeof(struct HMACMD5Context), GFP_KERNEL); + + if (pctxt == NULL) + return -ENOMEM; /* calculate md4 hash of password */ E_md4hash(ses->password, nt_hash); - crypto_shash_setkey(ses->server->ntlmssp.hmacmd5, nt_hash, - CIFS_NTHASH_SIZE); - - rc = crypto_shash_init(&ses->server->ntlmssp.sdeschmacmd5->shash); - if (rc) { - cERROR(1, "calc_ntlmv2_hash: could not init hmacmd5\n"); - return rc; - } + /* convert Domainname to unicode and uppercase */ + hmac_md5_init_limK_to_64(nt_hash, 16, pctxt); /* convert ses->userName to unicode and uppercase */ len = strlen(ses->userName); user = kmalloc(2 + (len * 2), GFP_KERNEL); - if (user == NULL) { - cERROR(1, "calc_ntlmv2_hash: user mem alloc failure\n"); - rc = -ENOMEM; + if (user == NULL) goto calc_exit_2; - } len = cifs_strtoUCS((__le16 *)user, ses->userName, len, nls_cp); UniStrupr(user); - - crypto_shash_update(&ses->server->ntlmssp.sdeschmacmd5->shash, - (char *)user, 2 * len); + hmac_md5_update((char *)user, 2*len, pctxt); /* convert ses->domainName to unicode and uppercase */ if (ses->domainName) { len = strlen(ses->domainName); domain = kmalloc(2 + (len * 2), GFP_KERNEL); - if (domain == NULL) { - cERROR(1, "calc_ntlmv2_hash: domain mem alloc failure"); - rc = -ENOMEM; + if (domain == NULL) goto calc_exit_1; - } len = cifs_strtoUCS((__le16 *)domain, ses->domainName, len, nls_cp); /* the following line was removed since it didn't work well @@ -359,292 +306,65 @@ static int calc_ntlmv2_hash(struct cifsSesInfo *ses, Maybe converting the domain name earlier makes sense */ /* UniStrupr(domain); */ - crypto_shash_update(&ses->server->ntlmssp.sdeschmacmd5->shash, - (char *)domain, 2 * len); + hmac_md5_update((char *)domain, 2*len, pctxt); kfree(domain); - } else if (ses->serverName) { - len = strlen(ses->serverName); - - server = kmalloc(2 + (len * 2), GFP_KERNEL); - if (server == NULL) { - cERROR(1, "calc_ntlmv2_hash: server mem alloc failure"); - rc = -ENOMEM; - goto calc_exit_1; - } - len = cifs_strtoUCS((__le16 *)server, ses->serverName, len, - nls_cp); - /* the following line was removed since it didn't work well - with lower cased domain name that passed as an option. - Maybe converting the domain name earlier makes sense */ - /* UniStrupr(domain); */ - - crypto_shash_update(&ses->server->ntlmssp.sdeschmacmd5->shash, - (char *)server, 2 * len); - - kfree(server); } - - rc = crypto_shash_final(&ses->server->ntlmssp.sdeschmacmd5->shash, - ses->server->ntlmv2_hash); - calc_exit_1: kfree(user); calc_exit_2: /* BB FIXME what about bytes 24 through 40 of the signing key? compare with the NTLM example */ + hmac_md5_final(ses->server->ntlmv2_hash, pctxt); + kfree(pctxt); return rc; } -static int -find_domain_name(struct cifsSesInfo *ses) -{ - int rc = 0; - unsigned int attrsize; - unsigned int type; - unsigned char *blobptr; - struct ntlmssp2_name *attrptr; - - if (ses->server->tiblob) { - blobptr = ses->server->tiblob; - attrptr = (struct ntlmssp2_name *) blobptr; - - while ((type = attrptr->type) != 0) { - blobptr += 2; /* advance attr type */ - attrsize = attrptr->length; - blobptr += 2; /* advance attr size */ - if (type == NTLMSSP_AV_NB_DOMAIN_NAME) { - if (!ses->domainName) { - ses->domainName = - kmalloc(attrptr->length + 1, - GFP_KERNEL); - if (!ses->domainName) - return -ENOMEM; - cifs_from_ucs2(ses->domainName, - (__le16 *)blobptr, - attrptr->length, - attrptr->length, - load_nls_default(), false); - } - } - blobptr += attrsize; /* advance attr value */ - attrptr = (struct ntlmssp2_name *) blobptr; - } - } else { - ses->server->tilen = 2 * sizeof(struct ntlmssp2_name); - ses->server->tiblob = kmalloc(ses->server->tilen, GFP_KERNEL); - if (!ses->server->tiblob) { - ses->server->tilen = 0; - cERROR(1, "Challenge target info allocation failure"); - return -ENOMEM; - } - memset(ses->server->tiblob, 0x0, ses->server->tilen); - attrptr = (struct ntlmssp2_name *) ses->server->tiblob; - attrptr->type = cpu_to_le16(NTLMSSP_DOMAIN_TYPE); - } - - return rc; -} - -static int -CalcNTLMv2_response(const struct TCP_Server_Info *server, - char *v2_session_response) -{ - int rc; - - if (!server->ntlmssp.sdeschmacmd5) { - cERROR(1, "calc_ntlmv2_hash: can't generate ntlmv2 hash\n"); - return -1; - } - - crypto_shash_setkey(server->ntlmssp.hmacmd5, server->ntlmv2_hash, - CIFS_HMAC_MD5_HASH_SIZE); - - rc = crypto_shash_init(&server->ntlmssp.sdeschmacmd5->shash); - if (rc) { - cERROR(1, "CalcNTLMv2_response: could not init hmacmd5"); - return rc; - } - - memcpy(v2_session_response + CIFS_SERVER_CHALLENGE_SIZE, - server->cryptKey, CIFS_SERVER_CHALLENGE_SIZE); - crypto_shash_update(&server->ntlmssp.sdeschmacmd5->shash, - v2_session_response + CIFS_SERVER_CHALLENGE_SIZE, - sizeof(struct ntlmv2_resp) - CIFS_SERVER_CHALLENGE_SIZE); - - if (server->tilen) - crypto_shash_update(&server->ntlmssp.sdeschmacmd5->shash, - server->tiblob, server->tilen); - - rc = crypto_shash_final(&server->ntlmssp.sdeschmacmd5->shash, - v2_session_response); - - return rc; -} - -int -setup_ntlmv2_rsp(struct cifsSesInfo *ses, char *resp_buf, +void setup_ntlmv2_rsp(struct cifsSesInfo *ses, char *resp_buf, const struct nls_table *nls_cp) { - int rc = 0; + int rc; struct ntlmv2_resp *buf = (struct ntlmv2_resp *)resp_buf; + struct HMACMD5Context context; buf->blob_signature = cpu_to_le32(0x00000101); buf->reserved = 0; buf->time = cpu_to_le64(cifs_UnixTimeToNT(CURRENT_TIME)); get_random_bytes(&buf->client_chal, sizeof(buf->client_chal)); buf->reserved2 = 0; - - if (!ses->domainName) { - rc = find_domain_name(ses); - if (rc) { - cERROR(1, "could not get domain/server name rc %d", rc); - return rc; - } - } + buf->names[0].type = cpu_to_le16(NTLMSSP_DOMAIN_TYPE); + buf->names[0].length = 0; + buf->names[1].type = 0; + buf->names[1].length = 0; /* calculate buf->ntlmv2_hash */ rc = calc_ntlmv2_hash(ses, nls_cp); - if (rc) { - cERROR(1, "could not get v2 hash rc %d", rc); - return rc; - } - rc = CalcNTLMv2_response(ses->server, resp_buf); - if (rc) { + if (rc) cERROR(1, "could not get v2 hash rc %d", rc); - return rc; - } - - if (!ses->server->ntlmssp.sdeschmacmd5) { - cERROR(1, "calc_ntlmv2_hash: can't generate ntlmv2 hash\n"); - return -1; - } - - crypto_shash_setkey(ses->server->ntlmssp.hmacmd5, - ses->server->ntlmv2_hash, CIFS_HMAC_MD5_HASH_SIZE); + CalcNTLMv2_response(ses, resp_buf); - rc = crypto_shash_init(&ses->server->ntlmssp.sdeschmacmd5->shash); - if (rc) { - cERROR(1, "setup_ntlmv2_rsp: could not init hmacmd5\n"); - return rc; - } + /* now calculate the MAC key for NTLMv2 */ + hmac_md5_init_limK_to_64(ses->server->ntlmv2_hash, 16, &context); + hmac_md5_update(resp_buf, 16, &context); + hmac_md5_final(ses->server->mac_signing_key.data.ntlmv2.key, &context); - crypto_shash_update(&ses->server->ntlmssp.sdeschmacmd5->shash, - resp_buf, CIFS_HMAC_MD5_HASH_SIZE); - - rc = crypto_shash_final(&ses->server->ntlmssp.sdeschmacmd5->shash, - ses->server->session_key.data.ntlmv2.key); - - memcpy(&ses->server->session_key.data.ntlmv2.resp, resp_buf, - sizeof(struct ntlmv2_resp)); - ses->server->session_key.len = 16 + sizeof(struct ntlmv2_resp); - - return rc; + memcpy(&ses->server->mac_signing_key.data.ntlmv2.resp, resp_buf, + sizeof(struct ntlmv2_resp)); + ses->server->mac_signing_key.len = 16 + sizeof(struct ntlmv2_resp); } -int -calc_seckey(struct TCP_Server_Info *server) -{ - int rc; - unsigned char sec_key[CIFS_NTLMV2_SESSKEY_SIZE]; - struct crypto_blkcipher *tfm_arc4; - struct scatterlist sgin, sgout; - struct blkcipher_desc desc; - - get_random_bytes(sec_key, CIFS_NTLMV2_SESSKEY_SIZE); - - tfm_arc4 = crypto_alloc_blkcipher("ecb(arc4)", - 0, CRYPTO_ALG_ASYNC); - if (!tfm_arc4 || IS_ERR(tfm_arc4)) { - cERROR(1, "could not allocate " "master crypto API arc4\n"); - return 1; - } - - desc.tfm = tfm_arc4; - - crypto_blkcipher_setkey(tfm_arc4, - server->session_key.data.ntlmv2.key, CIFS_CPHTXT_SIZE); - sg_init_one(&sgin, sec_key, CIFS_CPHTXT_SIZE); - sg_init_one(&sgout, server->ntlmssp.ciphertext, CIFS_CPHTXT_SIZE); - rc = crypto_blkcipher_encrypt(&desc, &sgout, &sgin, CIFS_CPHTXT_SIZE); - - if (!rc) - memcpy(server->session_key.data.ntlmv2.key, - sec_key, CIFS_NTLMV2_SESSKEY_SIZE); - - crypto_free_blkcipher(tfm_arc4); - - return 0; -} - -void -cifs_crypto_shash_release(struct TCP_Server_Info *server) -{ - if (server->ntlmssp.md5) - crypto_free_shash(server->ntlmssp.md5); - - if (server->ntlmssp.hmacmd5) - crypto_free_shash(server->ntlmssp.hmacmd5); - - kfree(server->ntlmssp.sdeschmacmd5); - - kfree(server->ntlmssp.sdescmd5); -} - -int -cifs_crypto_shash_allocate(struct TCP_Server_Info *server) +void CalcNTLMv2_response(const struct cifsSesInfo *ses, + char *v2_session_response) { - int rc; - unsigned int size; - - server->ntlmssp.hmacmd5 = crypto_alloc_shash("hmac(md5)", 0, 0); - if (!server->ntlmssp.hmacmd5 || - IS_ERR(server->ntlmssp.hmacmd5)) { - cERROR(1, "could not allocate crypto hmacmd5\n"); - return 1; - } - - server->ntlmssp.md5 = crypto_alloc_shash("md5", 0, 0); - if (!server->ntlmssp.md5 || IS_ERR(server->ntlmssp.md5)) { - cERROR(1, "could not allocate crypto md5\n"); - rc = 1; - goto cifs_crypto_shash_allocate_ret1; - } - - size = sizeof(struct shash_desc) + - crypto_shash_descsize(server->ntlmssp.hmacmd5); - server->ntlmssp.sdeschmacmd5 = kmalloc(size, GFP_KERNEL); - if (!server->ntlmssp.sdeschmacmd5) { - cERROR(1, "cifs_crypto_shash_allocate: can't alloc hmacmd5\n"); - rc = -ENOMEM; - goto cifs_crypto_shash_allocate_ret2; - } - server->ntlmssp.sdeschmacmd5->shash.tfm = server->ntlmssp.hmacmd5; - server->ntlmssp.sdeschmacmd5->shash.flags = 0x0; + struct HMACMD5Context context; + /* rest of v2 struct already generated */ + memcpy(v2_session_response + 8, ses->server->cryptKey, 8); + hmac_md5_init_limK_to_64(ses->server->ntlmv2_hash, 16, &context); + hmac_md5_update(v2_session_response+8, + sizeof(struct ntlmv2_resp) - 8, &context); - size = sizeof(struct shash_desc) + - crypto_shash_descsize(server->ntlmssp.md5); - server->ntlmssp.sdescmd5 = kmalloc(size, GFP_KERNEL); - if (!server->ntlmssp.sdescmd5) { - cERROR(1, "cifs_crypto_shash_allocate: can't alloc md5\n"); - rc = -ENOMEM; - goto cifs_crypto_shash_allocate_ret3; - } - server->ntlmssp.sdescmd5->shash.tfm = server->ntlmssp.md5; - server->ntlmssp.sdescmd5->shash.flags = 0x0; - - return 0; - -cifs_crypto_shash_allocate_ret3: - kfree(server->ntlmssp.sdeschmacmd5); - -cifs_crypto_shash_allocate_ret2: - crypto_free_shash(server->ntlmssp.md5); - -cifs_crypto_shash_allocate_ret1: - crypto_free_shash(server->ntlmssp.hmacmd5); - - return rc; + hmac_md5_final(v2_session_response, &context); +/* cifs_dump_mem("v2_sess_rsp: ", v2_session_response, 32); */ } diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index c9d0cfc086e..0cdfb8c32ac 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -25,9 +25,6 @@ #include <linux/workqueue.h> #include "cifs_fs_sb.h" #include "cifsacl.h" -#include <crypto/internal/hash.h> -#include <linux/scatterlist.h> - /* * The sizes of various internal tables and strings */ @@ -100,7 +97,7 @@ enum protocolEnum { /* Netbios frames protocol not supported at this time */ }; -struct session_key { +struct mac_key { unsigned int len; union { char ntlm[CIFS_SESS_KEY_SIZE + 16]; @@ -123,21 +120,6 @@ struct cifs_cred { struct cifs_ace *aces; }; -struct sdesc { - struct shash_desc shash; - char ctx[]; -}; - -struct ntlmssp_auth { - __u32 client_flags; - __u32 server_flags; - unsigned char ciphertext[CIFS_CPHTXT_SIZE]; - struct crypto_shash *hmacmd5; - struct crypto_shash *md5; - struct sdesc *sdeschmacmd5; - struct sdesc *sdescmd5; -}; - /* ***************************************************************** * Except the CIFS PDUs themselves all the @@ -200,14 +182,11 @@ struct TCP_Server_Info { /* 16th byte of RFC1001 workstation name is always null */ char workstation_RFC1001_name[RFC1001_NAME_LEN_WITH_NULL]; __u32 sequence_number; /* needed for CIFS PDU signature */ - struct session_key session_key; + struct mac_key mac_signing_key; char ntlmv2_hash[16]; unsigned long lstrp; /* when we got last response from this server */ u16 dialect; /* dialect index that server chose */ /* extended security flavors that server supports */ - unsigned int tilen; /* length of the target info blob */ - unsigned char *tiblob; /* target info blob in challenge response */ - struct ntlmssp_auth ntlmssp; /* various keys, ciphers, flags */ bool sec_kerberos; /* supports plain Kerberos */ bool sec_mskerberos; /* supports legacy MS Kerberos */ bool sec_kerberosu2u; /* supports U2U Kerberos */ diff --git a/fs/cifs/cifspdu.h b/fs/cifs/cifspdu.h index 320e0fd0ba7..14d036d8db1 100644 --- a/fs/cifs/cifspdu.h +++ b/fs/cifs/cifspdu.h @@ -134,12 +134,6 @@ * Size of the session key (crypto key encrypted with the password */ #define CIFS_SESS_KEY_SIZE (24) -#define CIFS_CLIENT_CHALLENGE_SIZE (8) -#define CIFS_SERVER_CHALLENGE_SIZE (8) -#define CIFS_HMAC_MD5_HASH_SIZE (16) -#define CIFS_CPHTXT_SIZE (16) -#define CIFS_NTLMV2_SESSKEY_SIZE (16) -#define CIFS_NTHASH_SIZE (16) /* * Maximum user name length @@ -669,6 +663,7 @@ struct ntlmv2_resp { __le64 time; __u64 client_chal; /* random */ __u32 reserved2; + struct ntlmssp2_name names[2]; /* array of name entries could follow ending in minimum 4 byte struct */ } __attribute__((packed)); diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index 1378d913384..1d60c655e3e 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -87,8 +87,9 @@ extern unsigned int smbCalcSize_LE(struct smb_hdr *ptr); extern int decode_negTokenInit(unsigned char *security_blob, int length, struct TCP_Server_Info *server); extern int cifs_convert_address(struct sockaddr *dst, const char *src, int len); +extern int cifs_set_port(struct sockaddr *addr, const unsigned short int port); extern int cifs_fill_sockaddr(struct sockaddr *dst, const char *src, int len, - unsigned short int port); + const unsigned short int port); extern int map_smb_to_linux_error(struct smb_hdr *smb, int logErr); extern void header_assemble(struct smb_hdr *, char /* command */ , const struct cifsTconInfo *, int /* length of @@ -361,15 +362,13 @@ extern int cifs_sign_smb(struct smb_hdr *, struct TCP_Server_Info *, __u32 *); extern int cifs_sign_smb2(struct kvec *iov, int n_vec, struct TCP_Server_Info *, __u32 *); extern int cifs_verify_signature(struct smb_hdr *, - struct TCP_Server_Info *server, + const struct mac_key *mac_key, __u32 expected_sequence_number); -extern int cifs_calculate_session_key(struct session_key *key, const char *rn, +extern int cifs_calculate_mac_key(struct mac_key *key, const char *rn, const char *pass); -extern int setup_ntlmv2_rsp(struct cifsSesInfo *, char *, +extern void CalcNTLMv2_response(const struct cifsSesInfo *, char *); +extern void setup_ntlmv2_rsp(struct cifsSesInfo *, char *, const struct nls_table *); -extern int cifs_crypto_shash_allocate(struct TCP_Server_Info *); -extern void cifs_crypto_shash_release(struct TCP_Server_Info *); -extern int calc_seckey(struct TCP_Server_Info *); #ifdef CONFIG_CIFS_WEAK_PW_HASH extern void calc_lanman_hash(const char *password, const char *cryptkey, bool encrypt, char *lnm_session_key); diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 4bda920d1f7..7e83b356cc9 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -232,7 +232,7 @@ static int small_smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, void **request_buf) { - int rc = 0; + int rc; rc = cifs_reconnect_tcon(tcon, smb_command); if (rc) @@ -250,7 +250,7 @@ small_smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, if (tcon != NULL) cifs_stats_inc(&tcon->num_smbs_sent); - return rc; + return 0; } int @@ -281,16 +281,9 @@ small_smb_init_no_tc(const int smb_command, const int wct, /* If the return code is zero, this function must fill in request_buf pointer */ static int -smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, - void **request_buf /* returned */ , - void **response_buf /* returned */ ) +__smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, + void **request_buf, void **response_buf) { - int rc = 0; - - rc = cifs_reconnect_tcon(tcon, smb_command); - if (rc) - return rc; - *request_buf = cifs_buf_get(); if (*request_buf == NULL) { /* BB should we add a retry in here if not a writepage? */ @@ -309,7 +302,31 @@ smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, if (tcon != NULL) cifs_stats_inc(&tcon->num_smbs_sent); - return rc; + return 0; +} + +/* If the return code is zero, this function must fill in request_buf pointer */ +static int +smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, + void **request_buf, void **response_buf) +{ + int rc; + + rc = cifs_reconnect_tcon(tcon, smb_command); + if (rc) + return rc; + + return __smb_init(smb_command, wct, tcon, request_buf, response_buf); +} + +static int +smb_init_no_reconnect(int smb_command, int wct, struct cifsTconInfo *tcon, + void **request_buf, void **response_buf) +{ + if (tcon->ses->need_reconnect || tcon->need_reconnect) + return -EHOSTDOWN; + + return __smb_init(smb_command, wct, tcon, request_buf, response_buf); } static int validate_t2(struct smb_t2_rsp *pSMB) @@ -604,14 +621,11 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) else rc = -EINVAL; - if (server->secType == Kerberos) { - if (!server->sec_kerberos && - !server->sec_mskerberos) - rc = -EOPNOTSUPP; - } else if (server->secType == RawNTLMSSP) { - if (!server->sec_ntlmssp) - rc = -EOPNOTSUPP; - } else + if (server->sec_kerberos || server->sec_mskerberos) + server->secType = Kerberos; + else if (server->sec_ntlmssp) + server->secType = RawNTLMSSP; + else rc = -EOPNOTSUPP; } } else @@ -4537,8 +4551,8 @@ CIFSSMBQFSUnixInfo(const int xid, struct cifsTconInfo *tcon) cFYI(1, "In QFSUnixInfo"); QFSUnixRetry: - rc = smb_init(SMB_COM_TRANSACTION2, 15, tcon, (void **) &pSMB, - (void **) &pSMBr); + rc = smb_init_no_reconnect(SMB_COM_TRANSACTION2, 15, tcon, + (void **) &pSMB, (void **) &pSMBr); if (rc) return rc; @@ -4607,8 +4621,8 @@ CIFSSMBSetFSUnixInfo(const int xid, struct cifsTconInfo *tcon, __u64 cap) cFYI(1, "In SETFSUnixInfo"); SETFSUnixRetry: /* BB switch to small buf init to save memory */ - rc = smb_init(SMB_COM_TRANSACTION2, 15, tcon, (void **) &pSMB, - (void **) &pSMBr); + rc = smb_init_no_reconnect(SMB_COM_TRANSACTION2, 15, tcon, + (void **) &pSMB, (void **) &pSMBr); if (rc) return rc; diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index ec0ea4a43bd..88c84a38bcc 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -400,7 +400,9 @@ incomplete_rcv: cFYI(1, "call to reconnect done"); csocket = server->ssocket; continue; - } else if ((length == -ERESTARTSYS) || (length == -EAGAIN)) { + } else if (length == -ERESTARTSYS || + length == -EAGAIN || + length == -EINTR) { msleep(1); /* minimum sleep to prevent looping allowing socket to clear and app threads to set tcpStatus CifsNeedReconnect if server hung */ @@ -414,18 +416,6 @@ incomplete_rcv: } else continue; } else if (length <= 0) { - if (server->tcpStatus == CifsNew) { - cFYI(1, "tcp session abend after SMBnegprot"); - /* some servers kill the TCP session rather than - returning an SMB negprot error, in which - case reconnecting here is not going to help, - and so simply return error to mount */ - break; - } - if (!try_to_freeze() && (length == -EINTR)) { - cFYI(1, "cifsd thread killed"); - break; - } cFYI(1, "Reconnect after unexpected peek error %d", length); cifs_reconnect(server); @@ -466,27 +456,19 @@ incomplete_rcv: an error on SMB negprot response */ cFYI(1, "Negative RFC1002 Session Response Error 0x%x)", pdu_length); - if (server->tcpStatus == CifsNew) { - /* if nack on negprot (rather than - ret of smb negprot error) reconnecting - not going to help, ret error to mount */ - break; - } else { - /* give server a second to - clean up before reconnect attempt */ - msleep(1000); - /* always try 445 first on reconnect - since we get NACK on some if we ever - connected to port 139 (the NACK is - since we do not begin with RFC1001 - session initialize frame) */ - server->addr.sockAddr.sin_port = - htons(CIFS_PORT); - cifs_reconnect(server); - csocket = server->ssocket; - wake_up(&server->response_q); - continue; - } + /* give server a second to clean up */ + msleep(1000); + /* always try 445 first on reconnect since we get NACK + * on some if we ever connected to port 139 (the NACK + * is since we do not begin with RFC1001 session + * initialize frame) + */ + cifs_set_port((struct sockaddr *) + &server->addr.sockAddr, CIFS_PORT); + cifs_reconnect(server); + csocket = server->ssocket; + wake_up(&server->response_q); + continue; } else if (temp != (char) 0) { cERROR(1, "Unknown RFC 1002 frame"); cifs_dump_mem(" Received Data: ", (char *)smb_buffer, @@ -522,8 +504,7 @@ incomplete_rcv: total_read += length) { length = kernel_recvmsg(csocket, &smb_msg, &iov, 1, pdu_length - total_read, 0); - if ((server->tcpStatus == CifsExiting) || - (length == -EINTR)) { + if (server->tcpStatus == CifsExiting) { /* then will exit */ reconnect = 2; break; @@ -534,8 +515,9 @@ incomplete_rcv: /* Now we will reread sock */ reconnect = 1; break; - } else if ((length == -ERESTARTSYS) || - (length == -EAGAIN)) { + } else if (length == -ERESTARTSYS || + length == -EAGAIN || + length == -EINTR) { msleep(1); /* minimum sleep to prevent looping, allowing socket to clear and app threads to set tcpStatus @@ -1708,7 +1690,6 @@ cifs_put_smb_ses(struct cifsSesInfo *ses) CIFSSMBLogoff(xid, ses); _FreeXid(xid); } - cifs_crypto_shash_release(server); sesInfoFree(ses); cifs_put_tcp_session(server); } @@ -1725,9 +1706,6 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) if (ses) { cFYI(1, "Existing smb sess found (status=%d)", ses->status); - /* existing SMB ses has a server reference already */ - cifs_put_tcp_session(server); - mutex_lock(&ses->session_mutex); rc = cifs_negotiate_protocol(xid, ses); if (rc) { @@ -1750,6 +1728,9 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) } } mutex_unlock(&ses->session_mutex); + + /* existing SMB ses has a server reference already */ + cifs_put_tcp_session(server); FreeXid(xid); return ses; } @@ -1788,23 +1769,13 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) ses->linux_uid = volume_info->linux_uid; ses->overrideSecFlg = volume_info->secFlg; - rc = cifs_crypto_shash_allocate(server); - if (rc) { - cERROR(1, "could not setup hash structures rc %d", rc); - goto get_ses_fail; - } - server->tilen = 0; - server->tiblob = NULL; - mutex_lock(&ses->session_mutex); rc = cifs_negotiate_protocol(xid, ses); if (!rc) rc = cifs_setup_session(xid, ses, volume_info->local_nls); mutex_unlock(&ses->session_mutex); - if (rc) { - cifs_crypto_shash_release(ses->server); + if (rc) goto get_ses_fail; - } /* success, put it on the list */ write_lock(&cifs_tcp_ses_lock); diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 86a164f08a7..53cce8cc222 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -801,6 +801,8 @@ retry_iget5_locked: inode->i_flags |= S_NOATIME | S_NOCMTIME; if (inode->i_state & I_NEW) { inode->i_ino = hash; + if (S_ISREG(inode->i_mode)) + inode->i_data.backing_dev_info = sb->s_bdi; #ifdef CONFIG_CIFS_FSCACHE /* initialize per-inode cache cookie pointer */ CIFS_I(inode)->fscache = NULL; @@ -1462,29 +1464,18 @@ int cifs_rename(struct inode *source_dir, struct dentry *source_dentry, { char *fromName = NULL; char *toName = NULL; - struct cifs_sb_info *cifs_sb_source; - struct cifs_sb_info *cifs_sb_target; + struct cifs_sb_info *cifs_sb; struct cifsTconInfo *tcon; FILE_UNIX_BASIC_INFO *info_buf_source = NULL; FILE_UNIX_BASIC_INFO *info_buf_target; int xid, rc, tmprc; - cifs_sb_target = CIFS_SB(target_dir->i_sb); - cifs_sb_source = CIFS_SB(source_dir->i_sb); - tcon = cifs_sb_source->tcon; + cifs_sb = CIFS_SB(source_dir->i_sb); + tcon = cifs_sb->tcon; xid = GetXid(); /* - * BB: this might be allowed if same server, but different share. - * Consider adding support for this - */ - if (tcon != cifs_sb_target->tcon) { - rc = -EXDEV; - goto cifs_rename_exit; - } - - /* * we already have the rename sem so we do not need to * grab it again here to protect the path integrity */ @@ -1519,17 +1510,16 @@ int cifs_rename(struct inode *source_dir, struct dentry *source_dentry, info_buf_target = info_buf_source + 1; tmprc = CIFSSMBUnixQPathInfo(xid, tcon, fromName, info_buf_source, - cifs_sb_source->local_nls, - cifs_sb_source->mnt_cifs_flags & + cifs_sb->local_nls, + cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); if (tmprc != 0) goto unlink_target; - tmprc = CIFSSMBUnixQPathInfo(xid, tcon, - toName, info_buf_target, - cifs_sb_target->local_nls, - /* remap based on source sb */ - cifs_sb_source->mnt_cifs_flags & + tmprc = CIFSSMBUnixQPathInfo(xid, tcon, toName, + info_buf_target, + cifs_sb->local_nls, + cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); if (tmprc == 0 && (info_buf_source->UniqueId == diff --git a/fs/cifs/netmisc.c b/fs/cifs/netmisc.c index f97851119e6..9aad47a2d62 100644 --- a/fs/cifs/netmisc.c +++ b/fs/cifs/netmisc.c @@ -206,26 +206,30 @@ cifs_convert_address(struct sockaddr *dst, const char *src, int len) } int -cifs_fill_sockaddr(struct sockaddr *dst, const char *src, int len, - const unsigned short int port) +cifs_set_port(struct sockaddr *addr, const unsigned short int port) { - if (!cifs_convert_address(dst, src, len)) - return 0; - - switch (dst->sa_family) { + switch (addr->sa_family) { case AF_INET: - ((struct sockaddr_in *)dst)->sin_port = htons(port); + ((struct sockaddr_in *)addr)->sin_port = htons(port); break; case AF_INET6: - ((struct sockaddr_in6 *)dst)->sin6_port = htons(port); + ((struct sockaddr_in6 *)addr)->sin6_port = htons(port); break; default: return 0; } - return 1; } +int +cifs_fill_sockaddr(struct sockaddr *dst, const char *src, int len, + const unsigned short int port) +{ + if (!cifs_convert_address(dst, src, len)) + return 0; + return cifs_set_port(dst, port); +} + /***************************************************************************** convert a NT status code to a dos class/code *****************************************************************************/ diff --git a/fs/cifs/ntlmssp.h b/fs/cifs/ntlmssp.h index 1db0f0746a5..49c9a4e7531 100644 --- a/fs/cifs/ntlmssp.h +++ b/fs/cifs/ntlmssp.h @@ -61,19 +61,6 @@ #define NTLMSSP_NEGOTIATE_KEY_XCH 0x40000000 #define NTLMSSP_NEGOTIATE_56 0x80000000 -/* Define AV Pair Field IDs */ -#define NTLMSSP_AV_EOL 0 -#define NTLMSSP_AV_NB_COMPUTER_NAME 1 -#define NTLMSSP_AV_NB_DOMAIN_NAME 2 -#define NTLMSSP_AV_DNS_COMPUTER_NAME 3 -#define NTLMSSP_AV_DNS_DOMAIN_NAME 4 -#define NTLMSSP_AV_DNS_TREE_NAME 5 -#define NTLMSSP_AV_FLAGS 6 -#define NTLMSSP_AV_TIMESTAMP 7 -#define NTLMSSP_AV_RESTRICTION 8 -#define NTLMSSP_AV_TARGET_NAME 9 -#define NTLMSSP_AV_CHANNEL_BINDINGS 10 - /* Although typedefs are not commonly used for structure definitions */ /* in the Linux kernel, in this particular case they are useful */ /* to more closely match the standards document for NTLMSSP from */ diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c index 795095f4eac..0a57cb7db5d 100644 --- a/fs/cifs/sess.c +++ b/fs/cifs/sess.c @@ -383,9 +383,6 @@ static int decode_ascii_ssetup(char **pbcc_area, int bleft, static int decode_ntlmssp_challenge(char *bcc_ptr, int blob_len, struct cifsSesInfo *ses) { - unsigned int tioffset; /* challeng message target info area */ - unsigned int tilen; /* challeng message target info area length */ - CHALLENGE_MESSAGE *pblob = (CHALLENGE_MESSAGE *)bcc_ptr; if (blob_len < sizeof(CHALLENGE_MESSAGE)) { @@ -408,20 +405,6 @@ static int decode_ntlmssp_challenge(char *bcc_ptr, int blob_len, /* BB spec says that if AvId field of MsvAvTimestamp is populated then we must set the MIC field of the AUTHENTICATE_MESSAGE */ - ses->server->ntlmssp.server_flags = le32_to_cpu(pblob->NegotiateFlags); - - tioffset = cpu_to_le16(pblob->TargetInfoArray.BufferOffset); - tilen = cpu_to_le16(pblob->TargetInfoArray.Length); - ses->server->tilen = tilen; - if (tilen) { - ses->server->tiblob = kmalloc(tilen, GFP_KERNEL); - if (!ses->server->tiblob) { - cERROR(1, "Challenge target info allocation failure"); - return -ENOMEM; - } - memcpy(ses->server->tiblob, bcc_ptr + tioffset, tilen); - } - return 0; } @@ -442,13 +425,12 @@ static void build_ntlmssp_negotiate_blob(unsigned char *pbuffer, /* BB is NTLMV2 session security format easier to use here? */ flags = NTLMSSP_NEGOTIATE_56 | NTLMSSP_REQUEST_TARGET | NTLMSSP_NEGOTIATE_128 | NTLMSSP_NEGOTIATE_UNICODE | - NTLMSSP_NEGOTIATE_NTLM; + NTLMSSP_NEGOTIATE_NT_ONLY | NTLMSSP_NEGOTIATE_NTLM; if (ses->server->secMode & - (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) { - flags |= NTLMSSP_NEGOTIATE_SIGN | - NTLMSSP_NEGOTIATE_KEY_XCH | - NTLMSSP_NEGOTIATE_EXTENDED_SEC; - } + (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) + flags |= NTLMSSP_NEGOTIATE_SIGN; + if (ses->server->secMode & SECMODE_SIGN_REQUIRED) + flags |= NTLMSSP_NEGOTIATE_ALWAYS_SIGN; sec_blob->NegotiateFlags |= cpu_to_le32(flags); @@ -469,12 +451,10 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, struct cifsSesInfo *ses, const struct nls_table *nls_cp, bool first) { - int rc; - unsigned int size; AUTHENTICATE_MESSAGE *sec_blob = (AUTHENTICATE_MESSAGE *)pbuffer; __u32 flags; unsigned char *tmp; - struct ntlmv2_resp ntlmv2_response = {}; + char ntlm_session_key[CIFS_SESS_KEY_SIZE]; memcpy(sec_blob->Signature, NTLMSSP_SIGNATURE, 8); sec_blob->MessageType = NtLmAuthenticate; @@ -497,25 +477,19 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, sec_blob->LmChallengeResponse.Length = 0; sec_blob->LmChallengeResponse.MaximumLength = 0; - sec_blob->NtChallengeResponse.BufferOffset = cpu_to_le32(tmp - pbuffer); - rc = setup_ntlmv2_rsp(ses, (char *)&ntlmv2_response, nls_cp); - if (rc) { - cERROR(1, "error rc: %d during ntlmssp ntlmv2 setup", rc); - goto setup_ntlmv2_ret; - } - size = sizeof(struct ntlmv2_resp); - memcpy(tmp, (char *)&ntlmv2_response, size); - tmp += size; - if (ses->server->tilen > 0) { - memcpy(tmp, ses->server->tiblob, ses->server->tilen); - tmp += ses->server->tilen; - } else - ses->server->tilen = 0; + /* calculate session key, BB what about adding similar ntlmv2 path? */ + SMBNTencrypt(ses->password, ses->server->cryptKey, ntlm_session_key); + if (first) + cifs_calculate_mac_key(&ses->server->mac_signing_key, + ntlm_session_key, ses->password); - sec_blob->NtChallengeResponse.Length = cpu_to_le16(size + - ses->server->tilen); + memcpy(tmp, ntlm_session_key, CIFS_SESS_KEY_SIZE); + sec_blob->NtChallengeResponse.BufferOffset = cpu_to_le32(tmp - pbuffer); + sec_blob->NtChallengeResponse.Length = cpu_to_le16(CIFS_SESS_KEY_SIZE); sec_blob->NtChallengeResponse.MaximumLength = - cpu_to_le16(size + ses->server->tilen); + cpu_to_le16(CIFS_SESS_KEY_SIZE); + + tmp += CIFS_SESS_KEY_SIZE; if (ses->domainName == NULL) { sec_blob->DomainName.BufferOffset = cpu_to_le32(tmp - pbuffer); @@ -527,6 +501,7 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, len = cifs_strtoUCS((__le16 *)tmp, ses->domainName, MAX_USERNAME_SIZE, nls_cp); len *= 2; /* unicode is 2 bytes each */ + len += 2; /* trailing null */ sec_blob->DomainName.BufferOffset = cpu_to_le32(tmp - pbuffer); sec_blob->DomainName.Length = cpu_to_le16(len); sec_blob->DomainName.MaximumLength = cpu_to_le16(len); @@ -543,6 +518,7 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, len = cifs_strtoUCS((__le16 *)tmp, ses->userName, MAX_USERNAME_SIZE, nls_cp); len *= 2; /* unicode is 2 bytes each */ + len += 2; /* trailing null */ sec_blob->UserName.BufferOffset = cpu_to_le32(tmp - pbuffer); sec_blob->UserName.Length = cpu_to_le16(len); sec_blob->UserName.MaximumLength = cpu_to_le16(len); @@ -554,26 +530,9 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, sec_blob->WorkstationName.MaximumLength = 0; tmp += 2; - if ((ses->server->ntlmssp.server_flags & NTLMSSP_NEGOTIATE_KEY_XCH) && - !calc_seckey(ses->server)) { - memcpy(tmp, ses->server->ntlmssp.ciphertext, CIFS_CPHTXT_SIZE); - sec_blob->SessionKey.BufferOffset = cpu_to_le32(tmp - pbuffer); - sec_blob->SessionKey.Length = cpu_to_le16(CIFS_CPHTXT_SIZE); - sec_blob->SessionKey.MaximumLength = - cpu_to_le16(CIFS_CPHTXT_SIZE); - tmp += CIFS_CPHTXT_SIZE; - } else { - sec_blob->SessionKey.BufferOffset = cpu_to_le32(tmp - pbuffer); - sec_blob->SessionKey.Length = 0; - sec_blob->SessionKey.MaximumLength = 0; - } - - ses->server->sequence_number = 0; - -setup_ntlmv2_ret: - if (ses->server->tilen > 0) - kfree(ses->server->tiblob); - + sec_blob->SessionKey.BufferOffset = cpu_to_le32(tmp - pbuffer); + sec_blob->SessionKey.Length = 0; + sec_blob->SessionKey.MaximumLength = 0; return tmp - pbuffer; } @@ -587,14 +546,15 @@ static void setup_ntlmssp_neg_req(SESSION_SETUP_ANDX *pSMB, return; } -static int setup_ntlmssp_auth_req(char *ntlmsspblob, +static int setup_ntlmssp_auth_req(SESSION_SETUP_ANDX *pSMB, struct cifsSesInfo *ses, const struct nls_table *nls, bool first_time) { int bloblen; - bloblen = build_ntlmssp_auth_blob(ntlmsspblob, ses, nls, + bloblen = build_ntlmssp_auth_blob(&pSMB->req.SecurityBlob[0], ses, nls, first_time); + pSMB->req.SecurityBlobLength = cpu_to_le16(bloblen); return bloblen; } @@ -730,7 +690,7 @@ ssetup_ntlmssp_authenticate: if (first_time) /* should this be moved into common code with similar ntlmv2 path? */ - cifs_calculate_session_key(&ses->server->session_key, + cifs_calculate_mac_key(&ses->server->mac_signing_key, ntlm_session_key, ses->password); /* copy session key */ @@ -769,21 +729,12 @@ ssetup_ntlmssp_authenticate: cpu_to_le16(sizeof(struct ntlmv2_resp)); /* calculate session key */ - rc = setup_ntlmv2_rsp(ses, v2_sess_key, nls_cp); - if (rc) { - kfree(v2_sess_key); - goto ssetup_exit; - } + setup_ntlmv2_rsp(ses, v2_sess_key, nls_cp); /* FIXME: calculate MAC key */ memcpy(bcc_ptr, (char *)v2_sess_key, sizeof(struct ntlmv2_resp)); bcc_ptr += sizeof(struct ntlmv2_resp); kfree(v2_sess_key); - if (ses->server->tilen > 0) { - memcpy(bcc_ptr, ses->server->tiblob, - ses->server->tilen); - bcc_ptr += ses->server->tilen; - } if (ses->capabilities & CAP_UNICODE) { if (iov[0].iov_len % 2) { *bcc_ptr = 0; @@ -814,15 +765,15 @@ ssetup_ntlmssp_authenticate: } /* bail out if key is too long */ if (msg->sesskey_len > - sizeof(ses->server->session_key.data.krb5)) { + sizeof(ses->server->mac_signing_key.data.krb5)) { cERROR(1, "Kerberos signing key too long (%u bytes)", msg->sesskey_len); rc = -EOVERFLOW; goto ssetup_exit; } if (first_time) { - ses->server->session_key.len = msg->sesskey_len; - memcpy(ses->server->session_key.data.krb5, + ses->server->mac_signing_key.len = msg->sesskey_len; + memcpy(ses->server->mac_signing_key.data.krb5, msg->data, msg->sesskey_len); } pSMB->req.hdr.Flags2 |= SMBFLG2_EXT_SEC; @@ -864,28 +815,12 @@ ssetup_ntlmssp_authenticate: if (phase == NtLmNegotiate) { setup_ntlmssp_neg_req(pSMB, ses); iov[1].iov_len = sizeof(NEGOTIATE_MESSAGE); - iov[1].iov_base = &pSMB->req.SecurityBlob[0]; } else if (phase == NtLmAuthenticate) { int blob_len; - char *ntlmsspblob; - - ntlmsspblob = kmalloc(5 * - sizeof(struct _AUTHENTICATE_MESSAGE), - GFP_KERNEL); - if (!ntlmsspblob) { - cERROR(1, "Can't allocate NTLMSSP"); - rc = -ENOMEM; - goto ssetup_exit; - } - - blob_len = setup_ntlmssp_auth_req(ntlmsspblob, - ses, - nls_cp, - first_time); + blob_len = setup_ntlmssp_auth_req(pSMB, ses, + nls_cp, + first_time); iov[1].iov_len = blob_len; - iov[1].iov_base = ntlmsspblob; - pSMB->req.SecurityBlobLength = - cpu_to_le16(blob_len); /* Make sure that we tell the server that we are using the uid that it just gave us back on the response (challenge) */ @@ -895,6 +830,7 @@ ssetup_ntlmssp_authenticate: rc = -ENOSYS; goto ssetup_exit; } + iov[1].iov_base = &pSMB->req.SecurityBlob[0]; /* unicode strings must be word aligned */ if ((iov[0].iov_len + iov[1].iov_len) % 2) { *bcc_ptr = 0; diff --git a/fs/cifs/transport.c b/fs/cifs/transport.c index e0588cdf4cc..82f78c4d697 100644 --- a/fs/cifs/transport.c +++ b/fs/cifs/transport.c @@ -543,7 +543,7 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, (ses->server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED))) { rc = cifs_verify_signature(midQ->resp_buf, - ses->server, + &ses->server->mac_signing_key, midQ->sequence_number+1); if (rc) { cERROR(1, "Unexpected SMB signature"); @@ -731,7 +731,7 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, (ses->server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED))) { rc = cifs_verify_signature(out_buf, - ses->server, + &ses->server->mac_signing_key, midQ->sequence_number+1); if (rc) { cERROR(1, "Unexpected SMB signature"); @@ -981,7 +981,7 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, (ses->server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED))) { rc = cifs_verify_signature(out_buf, - ses->server, + &ses->server->mac_signing_key, midQ->sequence_number+1); if (rc) { cERROR(1, "Unexpected SMB signature"); diff --git a/fs/coda/psdev.c b/fs/coda/psdev.c index de89645777c..116af7546cf 100644 --- a/fs/coda/psdev.c +++ b/fs/coda/psdev.c @@ -184,8 +184,8 @@ static ssize_t coda_psdev_write(struct file *file, const char __user *buf, } /* adjust outsize. is this useful ?? */ - req->uc_outSize = nbytes; - req->uc_flags |= REQ_WRITE; + req->uc_outSize = nbytes; + req->uc_flags |= CODA_REQ_WRITE; count = nbytes; /* Convert filedescriptor into a file handle */ diff --git a/fs/compat.c b/fs/compat.c index 718c7062aec..0644a154672 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -1153,7 +1153,7 @@ static ssize_t compat_do_readv_writev(int type, struct file *file, { compat_ssize_t tot_len; struct iovec iovstack[UIO_FASTIOV]; - struct iovec *iov; + struct iovec *iov = iovstack; ssize_t ret; io_fn_t fn; iov_fn_t fnv; diff --git a/fs/direct-io.c b/fs/direct-io.c index 51f270b479b..48d74c7391d 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -634,7 +634,7 @@ static int dio_send_cur_page(struct dio *dio) int ret = 0; if (dio->bio) { - loff_t cur_offset = dio->block_in_file << dio->blkbits; + loff_t cur_offset = dio->cur_page_fs_offset; loff_t bio_next_offset = dio->logical_offset_in_bio + dio->bio->bi_size; @@ -659,7 +659,7 @@ static int dio_send_cur_page(struct dio *dio) * Submit now if the underlying fs is about to perform a * metadata read */ - if (dio->boundary) + else if (dio->boundary) dio_bio_submit(dio); } diff --git a/fs/exec.c b/fs/exec.c index 2d945528274..828dd2461d6 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -376,6 +376,9 @@ static int count(const char __user * const __user * argv, int max) argv++; if (i++ >= max) return -E2BIG; + + if (fatal_signal_pending(current)) + return -ERESTARTNOHAND; cond_resched(); } } @@ -419,6 +422,12 @@ static int copy_strings(int argc, const char __user *const __user *argv, while (len > 0) { int offset, bytes_to_copy; + if (fatal_signal_pending(current)) { + ret = -ERESTARTNOHAND; + goto out; + } + cond_resched(); + offset = pos % PAGE_SIZE; if (offset == 0) offset = PAGE_SIZE; @@ -594,6 +603,11 @@ int setup_arg_pages(struct linux_binprm *bprm, #else stack_top = arch_align_stack(stack_top); stack_top = PAGE_ALIGN(stack_top); + + if (unlikely(stack_top < mmap_min_addr) || + unlikely(vma->vm_end - vma->vm_start >= stack_top - mmap_min_addr)) + return -ENOMEM; + stack_shift = vma->vm_end - stack_top; bprm->p -= stack_shift; diff --git a/fs/fcntl.c b/fs/fcntl.c index 6769fd0f35b..f8cc34f542c 100644 --- a/fs/fcntl.c +++ b/fs/fcntl.c @@ -769,11 +769,15 @@ EXPORT_SYMBOL(kill_fasync); static int __init fcntl_init(void) { - /* please add new bits here to ensure allocation uniqueness */ - BUILD_BUG_ON(19 - 1 /* for O_RDONLY being 0 */ != HWEIGHT32( + /* + * Please add new bits here to ensure allocation uniqueness. + * Exceptions: O_NONBLOCK is a two bit define on parisc; O_NDELAY + * is defined as O_NONBLOCK on some platforms and not on others. + */ + BUILD_BUG_ON(18 - 1 /* for O_RDONLY being 0 */ != HWEIGHT32( O_RDONLY | O_WRONLY | O_RDWR | O_CREAT | O_EXCL | O_NOCTTY | - O_TRUNC | O_APPEND | O_NONBLOCK | + O_TRUNC | O_APPEND | /* O_NONBLOCK | */ __O_SYNC | O_DSYNC | FASYNC | O_DIRECT | O_LARGEFILE | O_DIRECTORY | O_NOFOLLOW | O_NOATIME | O_CLOEXEC | diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 7d9d06ba184..ab38fef1c9a 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -52,8 +52,6 @@ struct wb_writeback_work { #define CREATE_TRACE_POINTS #include <trace/events/writeback.h> -#define inode_to_bdi(inode) ((inode)->i_mapping->backing_dev_info) - /* * We don't actually have pdflush, but this one is exported though /proc... */ @@ -71,6 +69,16 @@ int writeback_in_progress(struct backing_dev_info *bdi) return test_bit(BDI_writeback_running, &bdi->state); } +static inline struct backing_dev_info *inode_to_bdi(struct inode *inode) +{ + struct super_block *sb = inode->i_sb; + + if (strcmp(sb->s_type->name, "bdev") == 0) + return inode->i_mapping->backing_dev_info; + + return sb->s_bdi; +} + static void bdi_queue_work(struct backing_dev_info *bdi, struct wb_writeback_work *work) { @@ -808,7 +816,7 @@ int bdi_writeback_thread(void *data) wb->last_active = jiffies; set_current_state(TASK_INTERRUPTIBLE); - if (!list_empty(&bdi->work_list)) { + if (!list_empty(&bdi->work_list) || kthread_should_stop()) { __set_current_state(TASK_RUNNING); continue; } diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index d367af1514e..cde755cca56 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -1354,7 +1354,7 @@ static int fuse_retrieve(struct fuse_conn *fc, struct inode *inode, loff_t file_size; unsigned int num; unsigned int offset; - size_t total_len; + size_t total_len = 0; req = fuse_get_req(fc); if (IS_ERR(req)) diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c index cde1248a622..ac750bd31a6 100644 --- a/fs/gfs2/log.c +++ b/fs/gfs2/log.c @@ -932,7 +932,7 @@ int gfs2_logd(void *data) do { prepare_to_wait(&sdp->sd_logd_waitq, &wait, - TASK_UNINTERRUPTIBLE); + TASK_INTERRUPTIBLE); if (!gfs2_ail_flush_reqd(sdp) && !gfs2_jrnl_flush_reqd(sdp) && !kthread_should_stop()) diff --git a/fs/minix/namei.c b/fs/minix/namei.c index e20ee85955d..f3f3578393a 100644 --- a/fs/minix/namei.c +++ b/fs/minix/namei.c @@ -115,7 +115,7 @@ static int minix_mkdir(struct inode * dir, struct dentry *dentry, int mode) inode_inc_link_count(dir); - inode = minix_new_inode(dir, mode, &err); + inode = minix_new_inode(dir, S_IFDIR | mode, &err); if (!inode) goto out_dir; diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig index 6c2aad49d73..f7e13db613c 100644 --- a/fs/nfs/Kconfig +++ b/fs/nfs/Kconfig @@ -63,6 +63,7 @@ config NFS_V3_ACL config NFS_V4 bool "NFS client support for NFS version 4" depends on NFS_FS + select SUNRPC_GSS help This option enables support for version 4 of the NFS protocol (RFC 3530) in the kernel's NFS client. diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 4e7df2adb21..e7340729af8 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -275,7 +275,7 @@ static int nfs_sockaddr_match_ipaddr6(const struct sockaddr *sa1, sin1->sin6_scope_id != sin2->sin6_scope_id) return 0; - return ipv6_addr_equal(&sin1->sin6_addr, &sin1->sin6_addr); + return ipv6_addr_equal(&sin1->sin6_addr, &sin2->sin6_addr); } #else /* !defined(CONFIG_IPV6) && !defined(CONFIG_IPV6_MODULE) */ static int nfs_sockaddr_match_ipaddr6(const struct sockaddr *sa1, diff --git a/fs/nfs/file.c b/fs/nfs/file.c index eb51bd6201d..05bf3c0dc75 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -723,10 +723,6 @@ static int do_vfs_lock(struct file *file, struct file_lock *fl) default: BUG(); } - if (res < 0) - dprintk(KERN_WARNING "%s: VFS is out of sync with lock manager" - " - error %d!\n", - __func__, res); return res; } diff --git a/fs/nfs/super.c b/fs/nfs/super.c index ec3966e4706..f4cbf0c306c 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -431,7 +431,15 @@ static int nfs_statfs(struct dentry *dentry, struct kstatfs *buf) goto out_err; error = server->nfs_client->rpc_ops->statfs(server, fh, &res); + if (unlikely(error == -ESTALE)) { + struct dentry *pd_dentry; + pd_dentry = dget_parent(dentry); + if (pd_dentry != NULL) { + nfs_zap_caches(pd_dentry->d_inode); + dput(pd_dentry); + } + } nfs_free_fattr(res.fattr); if (error < 0) goto out_err; diff --git a/fs/nfsd/Kconfig b/fs/nfsd/Kconfig index 95932f523ae..4264377552e 100644 --- a/fs/nfsd/Kconfig +++ b/fs/nfsd/Kconfig @@ -69,6 +69,7 @@ config NFSD_V4 depends on NFSD && PROC_FS && EXPERIMENTAL select NFSD_V3 select FS_POSIX_ACL + select SUNRPC_GSS help This option enables support in your system's NFS server for version 4 of the NFS protocol (RFC 3530). diff --git a/fs/ocfs2/acl.c b/fs/ocfs2/acl.c index a76e0aa5cd3..391915093fe 100644 --- a/fs/ocfs2/acl.c +++ b/fs/ocfs2/acl.c @@ -209,7 +209,10 @@ static int ocfs2_acl_set_mode(struct inode *inode, struct buffer_head *di_bh, } inode->i_mode = new_mode; + inode->i_ctime = CURRENT_TIME; di->i_mode = cpu_to_le16(inode->i_mode); + di->i_ctime = cpu_to_le64(inode->i_ctime.tv_sec); + di->i_ctime_nsec = cpu_to_le32(inode->i_ctime.tv_nsec); ocfs2_journal_dirty(handle, di_bh); diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c index 41d5f1f92d5..52c7557f3e2 100644 --- a/fs/ocfs2/cluster/heartbeat.c +++ b/fs/ocfs2/cluster/heartbeat.c @@ -62,10 +62,51 @@ static unsigned long o2hb_live_node_bitmap[BITS_TO_LONGS(O2NM_MAX_NODES)]; static LIST_HEAD(o2hb_node_events); static DECLARE_WAIT_QUEUE_HEAD(o2hb_steady_queue); +/* + * In global heartbeat, we maintain a series of region bitmaps. + * - o2hb_region_bitmap allows us to limit the region number to max region. + * - o2hb_live_region_bitmap tracks live regions (seen steady iterations). + * - o2hb_quorum_region_bitmap tracks live regions that have seen all nodes + * heartbeat on it. + * - o2hb_failed_region_bitmap tracks the regions that have seen io timeouts. + */ +static unsigned long o2hb_region_bitmap[BITS_TO_LONGS(O2NM_MAX_REGIONS)]; +static unsigned long o2hb_live_region_bitmap[BITS_TO_LONGS(O2NM_MAX_REGIONS)]; +static unsigned long o2hb_quorum_region_bitmap[BITS_TO_LONGS(O2NM_MAX_REGIONS)]; +static unsigned long o2hb_failed_region_bitmap[BITS_TO_LONGS(O2NM_MAX_REGIONS)]; + +#define O2HB_DB_TYPE_LIVENODES 0 +#define O2HB_DB_TYPE_LIVEREGIONS 1 +#define O2HB_DB_TYPE_QUORUMREGIONS 2 +#define O2HB_DB_TYPE_FAILEDREGIONS 3 +#define O2HB_DB_TYPE_REGION_LIVENODES 4 +#define O2HB_DB_TYPE_REGION_NUMBER 5 +#define O2HB_DB_TYPE_REGION_ELAPSED_TIME 6 +struct o2hb_debug_buf { + int db_type; + int db_size; + int db_len; + void *db_data; +}; + +static struct o2hb_debug_buf *o2hb_db_livenodes; +static struct o2hb_debug_buf *o2hb_db_liveregions; +static struct o2hb_debug_buf *o2hb_db_quorumregions; +static struct o2hb_debug_buf *o2hb_db_failedregions; + #define O2HB_DEBUG_DIR "o2hb" #define O2HB_DEBUG_LIVENODES "livenodes" +#define O2HB_DEBUG_LIVEREGIONS "live_regions" +#define O2HB_DEBUG_QUORUMREGIONS "quorum_regions" +#define O2HB_DEBUG_FAILEDREGIONS "failed_regions" +#define O2HB_DEBUG_REGION_NUMBER "num" +#define O2HB_DEBUG_REGION_ELAPSED_TIME "elapsed_time_in_ms" + static struct dentry *o2hb_debug_dir; static struct dentry *o2hb_debug_livenodes; +static struct dentry *o2hb_debug_liveregions; +static struct dentry *o2hb_debug_quorumregions; +static struct dentry *o2hb_debug_failedregions; static LIST_HEAD(o2hb_all_regions); @@ -77,7 +118,19 @@ static struct o2hb_callback *hbcall_from_type(enum o2hb_callback_type type); #define O2HB_DEFAULT_BLOCK_BITS 9 +enum o2hb_heartbeat_modes { + O2HB_HEARTBEAT_LOCAL = 0, + O2HB_HEARTBEAT_GLOBAL, + O2HB_HEARTBEAT_NUM_MODES, +}; + +char *o2hb_heartbeat_mode_desc[O2HB_HEARTBEAT_NUM_MODES] = { + "local", /* O2HB_HEARTBEAT_LOCAL */ + "global", /* O2HB_HEARTBEAT_GLOBAL */ +}; + unsigned int o2hb_dead_threshold = O2HB_DEFAULT_DEAD_THRESHOLD; +unsigned int o2hb_heartbeat_mode = O2HB_HEARTBEAT_LOCAL; /* Only sets a new threshold if there are no active regions. * @@ -94,6 +147,22 @@ static void o2hb_dead_threshold_set(unsigned int threshold) } } +static int o2hb_global_hearbeat_mode_set(unsigned int hb_mode) +{ + int ret = -1; + + if (hb_mode < O2HB_HEARTBEAT_NUM_MODES) { + spin_lock(&o2hb_live_lock); + if (list_empty(&o2hb_all_regions)) { + o2hb_heartbeat_mode = hb_mode; + ret = 0; + } + spin_unlock(&o2hb_live_lock); + } + + return ret; +} + struct o2hb_node_event { struct list_head hn_item; enum o2hb_callback_type hn_event_type; @@ -135,6 +204,18 @@ struct o2hb_region { struct block_device *hr_bdev; struct o2hb_disk_slot *hr_slots; + /* live node map of this region */ + unsigned long hr_live_node_bitmap[BITS_TO_LONGS(O2NM_MAX_NODES)]; + unsigned int hr_region_num; + + struct dentry *hr_debug_dir; + struct dentry *hr_debug_livenodes; + struct dentry *hr_debug_regnum; + struct dentry *hr_debug_elapsed_time; + struct o2hb_debug_buf *hr_db_livenodes; + struct o2hb_debug_buf *hr_db_regnum; + struct o2hb_debug_buf *hr_db_elapsed_time; + /* let the person setting up hb wait for it to return until it * has reached a 'steady' state. This will be fixed when we have * a more complete api that doesn't lead to this sort of fragility. */ @@ -163,8 +244,19 @@ struct o2hb_bio_wait_ctxt { int wc_error; }; +static int o2hb_pop_count(void *map, int count) +{ + int i = -1, pop = 0; + + while ((i = find_next_bit(map, count, i + 1)) < count) + pop++; + return pop; +} + static void o2hb_write_timeout(struct work_struct *work) { + int failed, quorum; + unsigned long flags; struct o2hb_region *reg = container_of(work, struct o2hb_region, hr_write_timeout_work.work); @@ -172,6 +264,28 @@ static void o2hb_write_timeout(struct work_struct *work) mlog(ML_ERROR, "Heartbeat write timeout to device %s after %u " "milliseconds\n", reg->hr_dev_name, jiffies_to_msecs(jiffies - reg->hr_last_timeout_start)); + + if (o2hb_global_heartbeat_active()) { + spin_lock_irqsave(&o2hb_live_lock, flags); + if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap)) + set_bit(reg->hr_region_num, o2hb_failed_region_bitmap); + failed = o2hb_pop_count(&o2hb_failed_region_bitmap, + O2NM_MAX_REGIONS); + quorum = o2hb_pop_count(&o2hb_quorum_region_bitmap, + O2NM_MAX_REGIONS); + spin_unlock_irqrestore(&o2hb_live_lock, flags); + + mlog(ML_HEARTBEAT, "Number of regions %d, failed regions %d\n", + quorum, failed); + + /* + * Fence if the number of failed regions >= half the number + * of quorum regions + */ + if ((failed << 1) < quorum) + return; + } + o2quo_disk_timeout(); } @@ -180,6 +294,11 @@ static void o2hb_arm_write_timeout(struct o2hb_region *reg) mlog(ML_HEARTBEAT, "Queue write timeout for %u ms\n", O2HB_MAX_WRITE_TIMEOUT_MS); + if (o2hb_global_heartbeat_active()) { + spin_lock(&o2hb_live_lock); + clear_bit(reg->hr_region_num, o2hb_failed_region_bitmap); + spin_unlock(&o2hb_live_lock); + } cancel_delayed_work(®->hr_write_timeout_work); reg->hr_last_timeout_start = jiffies; schedule_delayed_work(®->hr_write_timeout_work, @@ -513,6 +632,8 @@ static void o2hb_queue_node_event(struct o2hb_node_event *event, { assert_spin_locked(&o2hb_live_lock); + BUG_ON((!node) && (type != O2HB_NODE_DOWN_CB)); + event->hn_event_type = type; event->hn_node = node; event->hn_node_num = node_num; @@ -554,6 +675,35 @@ static void o2hb_shutdown_slot(struct o2hb_disk_slot *slot) o2nm_node_put(node); } +static void o2hb_set_quorum_device(struct o2hb_region *reg, + struct o2hb_disk_slot *slot) +{ + assert_spin_locked(&o2hb_live_lock); + + if (!o2hb_global_heartbeat_active()) + return; + + if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap)) + return; + + /* + * A region can be added to the quorum only when it sees all + * live nodes heartbeat on it. In other words, the region has been + * added to all nodes. + */ + if (memcmp(reg->hr_live_node_bitmap, o2hb_live_node_bitmap, + sizeof(o2hb_live_node_bitmap))) + return; + + if (slot->ds_changed_samples < O2HB_LIVE_THRESHOLD) + return; + + printk(KERN_NOTICE "o2hb: Region %s is now a quorum device\n", + config_item_name(®->hr_item)); + + set_bit(reg->hr_region_num, o2hb_quorum_region_bitmap); +} + static int o2hb_check_slot(struct o2hb_region *reg, struct o2hb_disk_slot *slot) { @@ -565,14 +715,22 @@ static int o2hb_check_slot(struct o2hb_region *reg, u64 cputime; unsigned int dead_ms = o2hb_dead_threshold * O2HB_REGION_TIMEOUT_MS; unsigned int slot_dead_ms; + int tmp; memcpy(hb_block, slot->ds_raw_block, reg->hr_block_bytes); - /* Is this correct? Do we assume that the node doesn't exist - * if we're not configured for him? */ + /* + * If a node is no longer configured but is still in the livemap, we + * may need to clear that bit from the livemap. + */ node = o2nm_get_node_by_num(slot->ds_node_num); - if (!node) - return 0; + if (!node) { + spin_lock(&o2hb_live_lock); + tmp = test_bit(slot->ds_node_num, o2hb_live_node_bitmap); + spin_unlock(&o2hb_live_lock); + if (!tmp) + return 0; + } if (!o2hb_verify_crc(reg, hb_block)) { /* all paths from here will drop o2hb_live_lock for @@ -639,8 +797,12 @@ fire_callbacks: mlog(ML_HEARTBEAT, "Node %d (id 0x%llx) joined my region\n", slot->ds_node_num, (long long)slot->ds_last_generation); + set_bit(slot->ds_node_num, reg->hr_live_node_bitmap); + /* first on the list generates a callback */ if (list_empty(&o2hb_live_slots[slot->ds_node_num])) { + mlog(ML_HEARTBEAT, "o2hb: Add node %d to live nodes " + "bitmap\n", slot->ds_node_num); set_bit(slot->ds_node_num, o2hb_live_node_bitmap); o2hb_queue_node_event(&event, O2HB_NODE_UP_CB, node, @@ -684,13 +846,18 @@ fire_callbacks: mlog(ML_HEARTBEAT, "Node %d left my region\n", slot->ds_node_num); + clear_bit(slot->ds_node_num, reg->hr_live_node_bitmap); + /* last off the live_slot generates a callback */ list_del_init(&slot->ds_live_item); if (list_empty(&o2hb_live_slots[slot->ds_node_num])) { + mlog(ML_HEARTBEAT, "o2hb: Remove node %d from live " + "nodes bitmap\n", slot->ds_node_num); clear_bit(slot->ds_node_num, o2hb_live_node_bitmap); - o2hb_queue_node_event(&event, O2HB_NODE_DOWN_CB, node, - slot->ds_node_num); + /* node can be null */ + o2hb_queue_node_event(&event, O2HB_NODE_DOWN_CB, + node, slot->ds_node_num); changed = 1; } @@ -706,11 +873,14 @@ fire_callbacks: slot->ds_equal_samples = 0; } out: + o2hb_set_quorum_device(reg, slot); + spin_unlock(&o2hb_live_lock); o2hb_run_event_list(&event); - o2nm_node_put(node); + if (node) + o2nm_node_put(node); return changed; } @@ -737,6 +907,7 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg) { int i, ret, highest_node, change = 0; unsigned long configured_nodes[BITS_TO_LONGS(O2NM_MAX_NODES)]; + unsigned long live_node_bitmap[BITS_TO_LONGS(O2NM_MAX_NODES)]; struct o2hb_bio_wait_ctxt write_wc; ret = o2nm_configured_node_map(configured_nodes, @@ -746,6 +917,17 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg) return ret; } + /* + * If a node is not configured but is in the livemap, we still need + * to read the slot so as to be able to remove it from the livemap. + */ + o2hb_fill_node_map(live_node_bitmap, sizeof(live_node_bitmap)); + i = -1; + while ((i = find_next_bit(live_node_bitmap, + O2NM_MAX_NODES, i + 1)) < O2NM_MAX_NODES) { + set_bit(i, configured_nodes); + } + highest_node = o2hb_highest_node(configured_nodes, O2NM_MAX_NODES); if (highest_node >= O2NM_MAX_NODES) { mlog(ML_NOTICE, "ocfs2_heartbeat: no configured nodes found!\n"); @@ -917,21 +1099,59 @@ static int o2hb_thread(void *data) #ifdef CONFIG_DEBUG_FS static int o2hb_debug_open(struct inode *inode, struct file *file) { + struct o2hb_debug_buf *db = inode->i_private; + struct o2hb_region *reg; unsigned long map[BITS_TO_LONGS(O2NM_MAX_NODES)]; char *buf = NULL; int i = -1; int out = 0; + /* max_nodes should be the largest bitmap we pass here */ + BUG_ON(sizeof(map) < db->db_size); + buf = kmalloc(PAGE_SIZE, GFP_KERNEL); if (!buf) goto bail; - o2hb_fill_node_map(map, sizeof(map)); + switch (db->db_type) { + case O2HB_DB_TYPE_LIVENODES: + case O2HB_DB_TYPE_LIVEREGIONS: + case O2HB_DB_TYPE_QUORUMREGIONS: + case O2HB_DB_TYPE_FAILEDREGIONS: + spin_lock(&o2hb_live_lock); + memcpy(map, db->db_data, db->db_size); + spin_unlock(&o2hb_live_lock); + break; + + case O2HB_DB_TYPE_REGION_LIVENODES: + spin_lock(&o2hb_live_lock); + reg = (struct o2hb_region *)db->db_data; + memcpy(map, reg->hr_live_node_bitmap, db->db_size); + spin_unlock(&o2hb_live_lock); + break; + + case O2HB_DB_TYPE_REGION_NUMBER: + reg = (struct o2hb_region *)db->db_data; + out += snprintf(buf + out, PAGE_SIZE - out, "%d\n", + reg->hr_region_num); + goto done; + + case O2HB_DB_TYPE_REGION_ELAPSED_TIME: + reg = (struct o2hb_region *)db->db_data; + out += snprintf(buf + out, PAGE_SIZE - out, "%u\n", + jiffies_to_msecs(jiffies - + reg->hr_last_timeout_start)); + goto done; + + default: + goto done; + } - while ((i = find_next_bit(map, O2NM_MAX_NODES, i + 1)) < O2NM_MAX_NODES) + while ((i = find_next_bit(map, db->db_len, i + 1)) < db->db_len) out += snprintf(buf + out, PAGE_SIZE - out, "%d ", i); out += snprintf(buf + out, PAGE_SIZE - out, "\n"); +done: i_size_write(inode, out); file->private_data = buf; @@ -978,10 +1198,104 @@ static const struct file_operations o2hb_debug_fops = { void o2hb_exit(void) { - if (o2hb_debug_livenodes) - debugfs_remove(o2hb_debug_livenodes); - if (o2hb_debug_dir) - debugfs_remove(o2hb_debug_dir); + kfree(o2hb_db_livenodes); + kfree(o2hb_db_liveregions); + kfree(o2hb_db_quorumregions); + kfree(o2hb_db_failedregions); + debugfs_remove(o2hb_debug_failedregions); + debugfs_remove(o2hb_debug_quorumregions); + debugfs_remove(o2hb_debug_liveregions); + debugfs_remove(o2hb_debug_livenodes); + debugfs_remove(o2hb_debug_dir); +} + +static struct dentry *o2hb_debug_create(const char *name, struct dentry *dir, + struct o2hb_debug_buf **db, int db_len, + int type, int size, int len, void *data) +{ + *db = kmalloc(db_len, GFP_KERNEL); + if (!*db) + return NULL; + + (*db)->db_type = type; + (*db)->db_size = size; + (*db)->db_len = len; + (*db)->db_data = data; + + return debugfs_create_file(name, S_IFREG|S_IRUSR, dir, *db, + &o2hb_debug_fops); +} + +static int o2hb_debug_init(void) +{ + int ret = -ENOMEM; + + o2hb_debug_dir = debugfs_create_dir(O2HB_DEBUG_DIR, NULL); + if (!o2hb_debug_dir) { + mlog_errno(ret); + goto bail; + } + + o2hb_debug_livenodes = o2hb_debug_create(O2HB_DEBUG_LIVENODES, + o2hb_debug_dir, + &o2hb_db_livenodes, + sizeof(*o2hb_db_livenodes), + O2HB_DB_TYPE_LIVENODES, + sizeof(o2hb_live_node_bitmap), + O2NM_MAX_NODES, + o2hb_live_node_bitmap); + if (!o2hb_debug_livenodes) { + mlog_errno(ret); + goto bail; + } + + o2hb_debug_liveregions = o2hb_debug_create(O2HB_DEBUG_LIVEREGIONS, + o2hb_debug_dir, + &o2hb_db_liveregions, + sizeof(*o2hb_db_liveregions), + O2HB_DB_TYPE_LIVEREGIONS, + sizeof(o2hb_live_region_bitmap), + O2NM_MAX_REGIONS, + o2hb_live_region_bitmap); + if (!o2hb_debug_liveregions) { + mlog_errno(ret); + goto bail; + } + + o2hb_debug_quorumregions = + o2hb_debug_create(O2HB_DEBUG_QUORUMREGIONS, + o2hb_debug_dir, + &o2hb_db_quorumregions, + sizeof(*o2hb_db_quorumregions), + O2HB_DB_TYPE_QUORUMREGIONS, + sizeof(o2hb_quorum_region_bitmap), + O2NM_MAX_REGIONS, + o2hb_quorum_region_bitmap); + if (!o2hb_debug_quorumregions) { + mlog_errno(ret); + goto bail; + } + + o2hb_debug_failedregions = + o2hb_debug_create(O2HB_DEBUG_FAILEDREGIONS, + o2hb_debug_dir, + &o2hb_db_failedregions, + sizeof(*o2hb_db_failedregions), + O2HB_DB_TYPE_FAILEDREGIONS, + sizeof(o2hb_failed_region_bitmap), + O2NM_MAX_REGIONS, + o2hb_failed_region_bitmap); + if (!o2hb_debug_failedregions) { + mlog_errno(ret); + goto bail; + } + + ret = 0; +bail: + if (ret) + o2hb_exit(); + + return ret; } int o2hb_init(void) @@ -997,24 +1311,12 @@ int o2hb_init(void) INIT_LIST_HEAD(&o2hb_node_events); memset(o2hb_live_node_bitmap, 0, sizeof(o2hb_live_node_bitmap)); + memset(o2hb_region_bitmap, 0, sizeof(o2hb_region_bitmap)); + memset(o2hb_live_region_bitmap, 0, sizeof(o2hb_live_region_bitmap)); + memset(o2hb_quorum_region_bitmap, 0, sizeof(o2hb_quorum_region_bitmap)); + memset(o2hb_failed_region_bitmap, 0, sizeof(o2hb_failed_region_bitmap)); - o2hb_debug_dir = debugfs_create_dir(O2HB_DEBUG_DIR, NULL); - if (!o2hb_debug_dir) { - mlog_errno(-ENOMEM); - return -ENOMEM; - } - - o2hb_debug_livenodes = debugfs_create_file(O2HB_DEBUG_LIVENODES, - S_IFREG|S_IRUSR, - o2hb_debug_dir, NULL, - &o2hb_debug_fops); - if (!o2hb_debug_livenodes) { - mlog_errno(-ENOMEM); - debugfs_remove(o2hb_debug_dir); - return -ENOMEM; - } - - return 0; + return o2hb_debug_init(); } /* if we're already in a callback then we're already serialized by the sem */ @@ -1078,6 +1380,13 @@ static void o2hb_region_release(struct config_item *item) if (reg->hr_slots) kfree(reg->hr_slots); + kfree(reg->hr_db_regnum); + kfree(reg->hr_db_livenodes); + debugfs_remove(reg->hr_debug_livenodes); + debugfs_remove(reg->hr_debug_regnum); + debugfs_remove(reg->hr_debug_elapsed_time); + debugfs_remove(reg->hr_debug_dir); + spin_lock(&o2hb_live_lock); list_del(®->hr_all_item); spin_unlock(&o2hb_live_lock); @@ -1441,6 +1750,8 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg, /* Ok, we were woken. Make sure it wasn't by drop_item() */ spin_lock(&o2hb_live_lock); hb_task = reg->hr_task; + if (o2hb_global_heartbeat_active()) + set_bit(reg->hr_region_num, o2hb_live_region_bitmap); spin_unlock(&o2hb_live_lock); if (hb_task) @@ -1448,6 +1759,10 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg, else ret = -EIO; + if (hb_task && o2hb_global_heartbeat_active()) + printk(KERN_NOTICE "o2hb: Heartbeat started on region %s\n", + config_item_name(®->hr_item)); + out: if (filp) fput(filp); @@ -1586,21 +1901,94 @@ static struct o2hb_heartbeat_group *to_o2hb_heartbeat_group(struct config_group : NULL; } +static int o2hb_debug_region_init(struct o2hb_region *reg, struct dentry *dir) +{ + int ret = -ENOMEM; + + reg->hr_debug_dir = + debugfs_create_dir(config_item_name(®->hr_item), dir); + if (!reg->hr_debug_dir) { + mlog_errno(ret); + goto bail; + } + + reg->hr_debug_livenodes = + o2hb_debug_create(O2HB_DEBUG_LIVENODES, + reg->hr_debug_dir, + &(reg->hr_db_livenodes), + sizeof(*(reg->hr_db_livenodes)), + O2HB_DB_TYPE_REGION_LIVENODES, + sizeof(reg->hr_live_node_bitmap), + O2NM_MAX_NODES, reg); + if (!reg->hr_debug_livenodes) { + mlog_errno(ret); + goto bail; + } + + reg->hr_debug_regnum = + o2hb_debug_create(O2HB_DEBUG_REGION_NUMBER, + reg->hr_debug_dir, + &(reg->hr_db_regnum), + sizeof(*(reg->hr_db_regnum)), + O2HB_DB_TYPE_REGION_NUMBER, + 0, O2NM_MAX_NODES, reg); + if (!reg->hr_debug_regnum) { + mlog_errno(ret); + goto bail; + } + + reg->hr_debug_elapsed_time = + o2hb_debug_create(O2HB_DEBUG_REGION_ELAPSED_TIME, + reg->hr_debug_dir, + &(reg->hr_db_elapsed_time), + sizeof(*(reg->hr_db_elapsed_time)), + O2HB_DB_TYPE_REGION_ELAPSED_TIME, + 0, 0, reg); + if (!reg->hr_debug_elapsed_time) { + mlog_errno(ret); + goto bail; + } + + ret = 0; +bail: + return ret; +} + static struct config_item *o2hb_heartbeat_group_make_item(struct config_group *group, const char *name) { struct o2hb_region *reg = NULL; + int ret; reg = kzalloc(sizeof(struct o2hb_region), GFP_KERNEL); if (reg == NULL) return ERR_PTR(-ENOMEM); - config_item_init_type_name(®->hr_item, name, &o2hb_region_type); + if (strlen(name) > O2HB_MAX_REGION_NAME_LEN) + return ERR_PTR(-ENAMETOOLONG); spin_lock(&o2hb_live_lock); + reg->hr_region_num = 0; + if (o2hb_global_heartbeat_active()) { + reg->hr_region_num = find_first_zero_bit(o2hb_region_bitmap, + O2NM_MAX_REGIONS); + if (reg->hr_region_num >= O2NM_MAX_REGIONS) { + spin_unlock(&o2hb_live_lock); + return ERR_PTR(-EFBIG); + } + set_bit(reg->hr_region_num, o2hb_region_bitmap); + } list_add_tail(®->hr_all_item, &o2hb_all_regions); spin_unlock(&o2hb_live_lock); + config_item_init_type_name(®->hr_item, name, &o2hb_region_type); + + ret = o2hb_debug_region_init(reg, o2hb_debug_dir); + if (ret) { + config_item_put(®->hr_item); + return ERR_PTR(ret); + } + return ®->hr_item; } @@ -1612,6 +2000,10 @@ static void o2hb_heartbeat_group_drop_item(struct config_group *group, /* stop the thread when the user removes the region dir */ spin_lock(&o2hb_live_lock); + if (o2hb_global_heartbeat_active()) { + clear_bit(reg->hr_region_num, o2hb_region_bitmap); + clear_bit(reg->hr_region_num, o2hb_live_region_bitmap); + } hb_task = reg->hr_task; reg->hr_task = NULL; spin_unlock(&o2hb_live_lock); @@ -1628,6 +2020,9 @@ static void o2hb_heartbeat_group_drop_item(struct config_group *group, wake_up(&o2hb_steady_queue); } + if (o2hb_global_heartbeat_active()) + printk(KERN_NOTICE "o2hb: Heartbeat stopped on region %s\n", + config_item_name(®->hr_item)); config_item_put(item); } @@ -1688,6 +2083,41 @@ static ssize_t o2hb_heartbeat_group_threshold_store(struct o2hb_heartbeat_group return count; } +static +ssize_t o2hb_heartbeat_group_mode_show(struct o2hb_heartbeat_group *group, + char *page) +{ + return sprintf(page, "%s\n", + o2hb_heartbeat_mode_desc[o2hb_heartbeat_mode]); +} + +static +ssize_t o2hb_heartbeat_group_mode_store(struct o2hb_heartbeat_group *group, + const char *page, size_t count) +{ + unsigned int i; + int ret; + size_t len; + + len = (page[count - 1] == '\n') ? count - 1 : count; + if (!len) + return -EINVAL; + + for (i = 0; i < O2HB_HEARTBEAT_NUM_MODES; ++i) { + if (strnicmp(page, o2hb_heartbeat_mode_desc[i], len)) + continue; + + ret = o2hb_global_hearbeat_mode_set(i); + if (!ret) + printk(KERN_NOTICE "o2hb: Heartbeat mode set to %s\n", + o2hb_heartbeat_mode_desc[i]); + return count; + } + + return -EINVAL; + +} + static struct o2hb_heartbeat_group_attribute o2hb_heartbeat_group_attr_threshold = { .attr = { .ca_owner = THIS_MODULE, .ca_name = "dead_threshold", @@ -1696,8 +2126,17 @@ static struct o2hb_heartbeat_group_attribute o2hb_heartbeat_group_attr_threshold .store = o2hb_heartbeat_group_threshold_store, }; +static struct o2hb_heartbeat_group_attribute o2hb_heartbeat_group_attr_mode = { + .attr = { .ca_owner = THIS_MODULE, + .ca_name = "mode", + .ca_mode = S_IRUGO | S_IWUSR }, + .show = o2hb_heartbeat_group_mode_show, + .store = o2hb_heartbeat_group_mode_store, +}; + static struct configfs_attribute *o2hb_heartbeat_group_attrs[] = { &o2hb_heartbeat_group_attr_threshold.attr, + &o2hb_heartbeat_group_attr_mode.attr, NULL, }; @@ -1963,3 +2402,34 @@ void o2hb_stop_all_regions(void) spin_unlock(&o2hb_live_lock); } EXPORT_SYMBOL_GPL(o2hb_stop_all_regions); + +int o2hb_get_all_regions(char *region_uuids, u8 max_regions) +{ + struct o2hb_region *reg; + int numregs = 0; + char *p; + + spin_lock(&o2hb_live_lock); + + p = region_uuids; + list_for_each_entry(reg, &o2hb_all_regions, hr_all_item) { + mlog(0, "Region: %s\n", config_item_name(®->hr_item)); + if (numregs < max_regions) { + memcpy(p, config_item_name(®->hr_item), + O2HB_MAX_REGION_NAME_LEN); + p += O2HB_MAX_REGION_NAME_LEN; + } + numregs++; + } + + spin_unlock(&o2hb_live_lock); + + return numregs; +} +EXPORT_SYMBOL_GPL(o2hb_get_all_regions); + +int o2hb_global_heartbeat_active(void) +{ + return (o2hb_heartbeat_mode == O2HB_HEARTBEAT_GLOBAL); +} +EXPORT_SYMBOL(o2hb_global_heartbeat_active); diff --git a/fs/ocfs2/cluster/heartbeat.h b/fs/ocfs2/cluster/heartbeat.h index 2f1649253b4..00ad8e8fea5 100644 --- a/fs/ocfs2/cluster/heartbeat.h +++ b/fs/ocfs2/cluster/heartbeat.h @@ -31,6 +31,8 @@ #define O2HB_REGION_TIMEOUT_MS 2000 +#define O2HB_MAX_REGION_NAME_LEN 32 + /* number of changes to be seen as live */ #define O2HB_LIVE_THRESHOLD 2 /* number of equal samples to be seen as dead */ @@ -81,5 +83,7 @@ int o2hb_check_node_heartbeating(u8 node_num); int o2hb_check_node_heartbeating_from_callback(u8 node_num); int o2hb_check_local_node_heartbeating(void); void o2hb_stop_all_regions(void); +int o2hb_get_all_regions(char *region_uuids, u8 numregions); +int o2hb_global_heartbeat_active(void); #endif /* O2CLUSTER_HEARTBEAT_H */ diff --git a/fs/ocfs2/cluster/masklog.h b/fs/ocfs2/cluster/masklog.h index fd96e2a2fa5..ea2ed9f56c9 100644 --- a/fs/ocfs2/cluster/masklog.h +++ b/fs/ocfs2/cluster/masklog.h @@ -119,7 +119,8 @@ #define ML_ERROR 0x0000000100000000ULL /* sent to KERN_ERR */ #define ML_NOTICE 0x0000000200000000ULL /* setn to KERN_NOTICE */ #define ML_KTHREAD 0x0000000400000000ULL /* kernel thread activity */ -#define ML_RESERVATIONS 0x0000000800000000ULL /* ocfs2 alloc reservations */ +#define ML_RESERVATIONS 0x0000000800000000ULL /* ocfs2 alloc reservations */ +#define ML_CLUSTER 0x0000001000000000ULL /* cluster stack */ #define MLOG_INITIAL_AND_MASK (ML_ERROR|ML_NOTICE) #define MLOG_INITIAL_NOT_MASK (ML_ENTRY|ML_EXIT) diff --git a/fs/ocfs2/cluster/nodemanager.c b/fs/ocfs2/cluster/nodemanager.c index ed0c9f367fe..bb240647ca5 100644 --- a/fs/ocfs2/cluster/nodemanager.c +++ b/fs/ocfs2/cluster/nodemanager.c @@ -711,6 +711,8 @@ static struct config_item *o2nm_node_group_make_item(struct config_group *group, config_item_init_type_name(&node->nd_item, name, &o2nm_node_type); spin_lock_init(&node->nd_lock); + mlog(ML_CLUSTER, "o2nm: Registering node %s\n", name); + return &node->nd_item; } @@ -744,6 +746,9 @@ static void o2nm_node_group_drop_item(struct config_group *group, } write_unlock(&cluster->cl_nodes_lock); + mlog(ML_CLUSTER, "o2nm: Unregistered node %s\n", + config_item_name(&node->nd_item)); + config_item_put(item); } diff --git a/fs/ocfs2/cluster/ocfs2_nodemanager.h b/fs/ocfs2/cluster/ocfs2_nodemanager.h index 5b9854bad57..49b594325be 100644 --- a/fs/ocfs2/cluster/ocfs2_nodemanager.h +++ b/fs/ocfs2/cluster/ocfs2_nodemanager.h @@ -36,4 +36,10 @@ /* host name, group name, cluster name all 64 bytes */ #define O2NM_MAX_NAME_LEN 64 // __NEW_UTS_LEN +/* + * Maximum number of global heartbeat regions allowed. + * **CAUTION** Changing this number will break dlm compatibility. + */ +#define O2NM_MAX_REGIONS 32 + #endif /* _OCFS2_NODEMANAGER_H */ diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index 1361997cf20..9aa426e4212 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -977,7 +977,7 @@ static int o2net_tx_can_proceed(struct o2net_node *nn, int o2net_send_message_vec(u32 msg_type, u32 key, struct kvec *caller_vec, size_t caller_veclen, u8 target_node, int *status) { - int ret; + int ret = 0; struct o2net_msg *msg = NULL; size_t veclen, caller_bytes = 0; struct kvec *vec = NULL; @@ -1696,6 +1696,9 @@ static void o2net_hb_node_down_cb(struct o2nm_node *node, int node_num, { o2quo_hb_down(node_num); + if (!node) + return; + if (node_num != o2nm_this_node()) o2net_disconnect_node(node); @@ -1709,6 +1712,8 @@ static void o2net_hb_node_up_cb(struct o2nm_node *node, int node_num, o2quo_hb_up(node_num); + BUG_ON(!node); + /* ensure an immediate connect attempt */ nn->nn_last_connect_attempt = jiffies - (msecs_to_jiffies(o2net_reconnect_delay()) + 1); diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c index f04ebcfffc4..c49f6de0e7a 100644 --- a/fs/ocfs2/dir.c +++ b/fs/ocfs2/dir.c @@ -3931,6 +3931,15 @@ static int ocfs2_dx_dir_rebalance(struct ocfs2_super *osb, struct inode *dir, goto out_commit; } + cpos = split_hash; + ret = ocfs2_dx_dir_new_cluster(dir, &et, cpos, handle, + data_ac, meta_ac, new_dx_leaves, + num_dx_leaves); + if (ret) { + mlog_errno(ret); + goto out_commit; + } + for (i = 0; i < num_dx_leaves; i++) { ret = ocfs2_journal_access_dl(handle, INODE_CACHE(dir), orig_dx_leaves[i], @@ -3939,15 +3948,14 @@ static int ocfs2_dx_dir_rebalance(struct ocfs2_super *osb, struct inode *dir, mlog_errno(ret); goto out_commit; } - } - cpos = split_hash; - ret = ocfs2_dx_dir_new_cluster(dir, &et, cpos, handle, - data_ac, meta_ac, new_dx_leaves, - num_dx_leaves); - if (ret) { - mlog_errno(ret); - goto out_commit; + ret = ocfs2_journal_access_dl(handle, INODE_CACHE(dir), + new_dx_leaves[i], + OCFS2_JOURNAL_ACCESS_WRITE); + if (ret) { + mlog_errno(ret); + goto out_commit; + } } ocfs2_dx_dir_transfer_leaf(dir, split_hash, handle, tmp_dx_leaf, diff --git a/fs/ocfs2/dlm/dlmcommon.h b/fs/ocfs2/dlm/dlmcommon.h index 4b6ae2c13b4..b36d0bf77a5 100644 --- a/fs/ocfs2/dlm/dlmcommon.h +++ b/fs/ocfs2/dlm/dlmcommon.h @@ -445,7 +445,9 @@ enum { DLM_LOCK_REQUEST_MSG, /* 515 */ DLM_RECO_DATA_DONE_MSG, /* 516 */ DLM_BEGIN_RECO_MSG, /* 517 */ - DLM_FINALIZE_RECO_MSG /* 518 */ + DLM_FINALIZE_RECO_MSG, /* 518 */ + DLM_QUERY_REGION, /* 519 */ + DLM_QUERY_NODEINFO, /* 520 */ }; struct dlm_reco_node_data @@ -727,6 +729,31 @@ struct dlm_cancel_join u8 domain[O2NM_MAX_NAME_LEN]; }; +struct dlm_query_region { + u8 qr_node; + u8 qr_numregions; + u8 qr_namelen; + u8 pad1; + u8 qr_domain[O2NM_MAX_NAME_LEN]; + u8 qr_regions[O2HB_MAX_REGION_NAME_LEN * O2NM_MAX_REGIONS]; +}; + +struct dlm_node_info { + u8 ni_nodenum; + u8 pad1; + u16 ni_ipv4_port; + u32 ni_ipv4_address; +}; + +struct dlm_query_nodeinfo { + u8 qn_nodenum; + u8 qn_numnodes; + u8 qn_namelen; + u8 pad1; + u8 qn_domain[O2NM_MAX_NAME_LEN]; + struct dlm_node_info qn_nodes[O2NM_MAX_NODES]; +}; + struct dlm_exit_domain { u8 node_idx; @@ -1030,6 +1057,7 @@ int dlm_drop_lockres_ref(struct dlm_ctxt *dlm, struct dlm_lock_resource *res); void dlm_clean_master_list(struct dlm_ctxt *dlm, u8 dead_node); +void dlm_force_free_mles(struct dlm_ctxt *dlm); int dlm_lock_basts_flushed(struct dlm_ctxt *dlm, struct dlm_lock *lock); int __dlm_lockres_has_locks(struct dlm_lock_resource *res); int __dlm_lockres_unused(struct dlm_lock_resource *res); diff --git a/fs/ocfs2/dlm/dlmdebug.c b/fs/ocfs2/dlm/dlmdebug.c index 26ff2e185b1..272ec8631a5 100644 --- a/fs/ocfs2/dlm/dlmdebug.c +++ b/fs/ocfs2/dlm/dlmdebug.c @@ -636,8 +636,14 @@ static void *lockres_seq_start(struct seq_file *m, loff_t *pos) spin_lock(&dlm->track_lock); if (oldres) track_list = &oldres->tracking; - else + else { track_list = &dlm->tracking_list; + if (list_empty(track_list)) { + dl = NULL; + spin_unlock(&dlm->track_lock); + goto bail; + } + } list_for_each_entry(res, track_list, tracking) { if (&res->tracking == &dlm->tracking_list) @@ -660,6 +666,7 @@ static void *lockres_seq_start(struct seq_file *m, loff_t *pos) } else dl = NULL; +bail: /* passed to seq_show */ return dl; } @@ -775,7 +782,9 @@ static int debug_state_print(struct dlm_ctxt *dlm, struct debug_buffer *db) /* Domain: xxxxxxxxxx Key: 0xdfbac769 */ out += snprintf(db->buf + out, db->len - out, - "Domain: %s Key: 0x%08x\n", dlm->name, dlm->key); + "Domain: %s Key: 0x%08x Protocol: %d.%d\n", + dlm->name, dlm->key, dlm->dlm_locking_proto.pv_major, + dlm->dlm_locking_proto.pv_minor); /* Thread Pid: xxx Node: xxx State: xxxxx */ out += snprintf(db->buf + out, db->len - out, diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c index 153abb5abef..58a93b95373 100644 --- a/fs/ocfs2/dlm/dlmdomain.c +++ b/fs/ocfs2/dlm/dlmdomain.c @@ -128,10 +128,14 @@ static DECLARE_WAIT_QUEUE_HEAD(dlm_domain_events); * will have a negotiated version with the same major number and a minor * number equal or smaller. The dlm_ctxt->dlm_locking_proto field should * be used to determine what a running domain is actually using. + * + * New in version 1.1: + * - Message DLM_QUERY_REGION added to support global heartbeat + * - Message DLM_QUERY_NODEINFO added to allow online node removes */ static const struct dlm_protocol_version dlm_protocol = { .pv_major = 1, - .pv_minor = 0, + .pv_minor = 1, }; #define DLM_DOMAIN_BACKOFF_MS 200 @@ -142,6 +146,8 @@ static int dlm_assert_joined_handler(struct o2net_msg *msg, u32 len, void *data, void **ret_data); static int dlm_cancel_join_handler(struct o2net_msg *msg, u32 len, void *data, void **ret_data); +static int dlm_query_region_handler(struct o2net_msg *msg, u32 len, + void *data, void **ret_data); static int dlm_exit_domain_handler(struct o2net_msg *msg, u32 len, void *data, void **ret_data); static int dlm_protocol_compare(struct dlm_protocol_version *existing, @@ -693,6 +699,7 @@ void dlm_unregister_domain(struct dlm_ctxt *dlm) dlm_mark_domain_leaving(dlm); dlm_leave_domain(dlm); + dlm_force_free_mles(dlm); dlm_complete_dlm_shutdown(dlm); } dlm_put(dlm); @@ -920,6 +927,370 @@ static int dlm_assert_joined_handler(struct o2net_msg *msg, u32 len, void *data, return 0; } +static int dlm_match_regions(struct dlm_ctxt *dlm, + struct dlm_query_region *qr) +{ + char *local = NULL, *remote = qr->qr_regions; + char *l, *r; + int localnr, i, j, foundit; + int status = 0; + + if (!o2hb_global_heartbeat_active()) { + if (qr->qr_numregions) { + mlog(ML_ERROR, "Domain %s: Joining node %d has global " + "heartbeat enabled but local node %d does not\n", + qr->qr_domain, qr->qr_node, dlm->node_num); + status = -EINVAL; + } + goto bail; + } + + if (o2hb_global_heartbeat_active() && !qr->qr_numregions) { + mlog(ML_ERROR, "Domain %s: Local node %d has global " + "heartbeat enabled but joining node %d does not\n", + qr->qr_domain, dlm->node_num, qr->qr_node); + status = -EINVAL; + goto bail; + } + + r = remote; + for (i = 0; i < qr->qr_numregions; ++i) { + mlog(0, "Region %.*s\n", O2HB_MAX_REGION_NAME_LEN, r); + r += O2HB_MAX_REGION_NAME_LEN; + } + + local = kmalloc(sizeof(qr->qr_regions), GFP_KERNEL); + if (!local) { + status = -ENOMEM; + goto bail; + } + + localnr = o2hb_get_all_regions(local, O2NM_MAX_REGIONS); + + /* compare local regions with remote */ + l = local; + for (i = 0; i < localnr; ++i) { + foundit = 0; + r = remote; + for (j = 0; j <= qr->qr_numregions; ++j) { + if (!memcmp(l, r, O2HB_MAX_REGION_NAME_LEN)) { + foundit = 1; + break; + } + r += O2HB_MAX_REGION_NAME_LEN; + } + if (!foundit) { + status = -EINVAL; + mlog(ML_ERROR, "Domain %s: Region '%.*s' registered " + "in local node %d but not in joining node %d\n", + qr->qr_domain, O2HB_MAX_REGION_NAME_LEN, l, + dlm->node_num, qr->qr_node); + goto bail; + } + l += O2HB_MAX_REGION_NAME_LEN; + } + + /* compare remote with local regions */ + r = remote; + for (i = 0; i < qr->qr_numregions; ++i) { + foundit = 0; + l = local; + for (j = 0; j < localnr; ++j) { + if (!memcmp(r, l, O2HB_MAX_REGION_NAME_LEN)) { + foundit = 1; + break; + } + l += O2HB_MAX_REGION_NAME_LEN; + } + if (!foundit) { + status = -EINVAL; + mlog(ML_ERROR, "Domain %s: Region '%.*s' registered " + "in joining node %d but not in local node %d\n", + qr->qr_domain, O2HB_MAX_REGION_NAME_LEN, r, + qr->qr_node, dlm->node_num); + goto bail; + } + r += O2HB_MAX_REGION_NAME_LEN; + } + +bail: + kfree(local); + + return status; +} + +static int dlm_send_regions(struct dlm_ctxt *dlm, unsigned long *node_map) +{ + struct dlm_query_region *qr = NULL; + int status, ret = 0, i; + char *p; + + if (find_next_bit(node_map, O2NM_MAX_NODES, 0) >= O2NM_MAX_NODES) + goto bail; + + qr = kzalloc(sizeof(struct dlm_query_region), GFP_KERNEL); + if (!qr) { + ret = -ENOMEM; + mlog_errno(ret); + goto bail; + } + + qr->qr_node = dlm->node_num; + qr->qr_namelen = strlen(dlm->name); + memcpy(qr->qr_domain, dlm->name, qr->qr_namelen); + /* if local hb, the numregions will be zero */ + if (o2hb_global_heartbeat_active()) + qr->qr_numregions = o2hb_get_all_regions(qr->qr_regions, + O2NM_MAX_REGIONS); + + p = qr->qr_regions; + for (i = 0; i < qr->qr_numregions; ++i, p += O2HB_MAX_REGION_NAME_LEN) + mlog(0, "Region %.*s\n", O2HB_MAX_REGION_NAME_LEN, p); + + i = -1; + while ((i = find_next_bit(node_map, O2NM_MAX_NODES, + i + 1)) < O2NM_MAX_NODES) { + if (i == dlm->node_num) + continue; + + mlog(0, "Sending regions to node %d\n", i); + + ret = o2net_send_message(DLM_QUERY_REGION, DLM_MOD_KEY, qr, + sizeof(struct dlm_query_region), + i, &status); + if (ret >= 0) + ret = status; + if (ret) { + mlog(ML_ERROR, "Region mismatch %d, node %d\n", + ret, i); + break; + } + } + +bail: + kfree(qr); + return ret; +} + +static int dlm_query_region_handler(struct o2net_msg *msg, u32 len, + void *data, void **ret_data) +{ + struct dlm_query_region *qr; + struct dlm_ctxt *dlm = NULL; + int status = 0; + int locked = 0; + + qr = (struct dlm_query_region *) msg->buf; + + mlog(0, "Node %u queries hb regions on domain %s\n", qr->qr_node, + qr->qr_domain); + + status = -EINVAL; + + spin_lock(&dlm_domain_lock); + dlm = __dlm_lookup_domain_full(qr->qr_domain, qr->qr_namelen); + if (!dlm) { + mlog(ML_ERROR, "Node %d queried hb regions on domain %s " + "before join domain\n", qr->qr_node, qr->qr_domain); + goto bail; + } + + spin_lock(&dlm->spinlock); + locked = 1; + if (dlm->joining_node != qr->qr_node) { + mlog(ML_ERROR, "Node %d queried hb regions on domain %s " + "but joining node is %d\n", qr->qr_node, qr->qr_domain, + dlm->joining_node); + goto bail; + } + + /* Support for global heartbeat was added in 1.1 */ + if (dlm->dlm_locking_proto.pv_major == 1 && + dlm->dlm_locking_proto.pv_minor == 0) { + mlog(ML_ERROR, "Node %d queried hb regions on domain %s " + "but active dlm protocol is %d.%d\n", qr->qr_node, + qr->qr_domain, dlm->dlm_locking_proto.pv_major, + dlm->dlm_locking_proto.pv_minor); + goto bail; + } + + status = dlm_match_regions(dlm, qr); + +bail: + if (locked) + spin_unlock(&dlm->spinlock); + spin_unlock(&dlm_domain_lock); + + return status; +} + +static int dlm_match_nodes(struct dlm_ctxt *dlm, struct dlm_query_nodeinfo *qn) +{ + struct o2nm_node *local; + struct dlm_node_info *remote; + int i, j; + int status = 0; + + for (j = 0; j < qn->qn_numnodes; ++j) + mlog(0, "Node %3d, %pI4:%u\n", qn->qn_nodes[j].ni_nodenum, + &(qn->qn_nodes[j].ni_ipv4_address), + ntohs(qn->qn_nodes[j].ni_ipv4_port)); + + for (i = 0; i < O2NM_MAX_NODES && !status; ++i) { + local = o2nm_get_node_by_num(i); + remote = NULL; + for (j = 0; j < qn->qn_numnodes; ++j) { + if (qn->qn_nodes[j].ni_nodenum == i) { + remote = &(qn->qn_nodes[j]); + break; + } + } + + if (!local && !remote) + continue; + + if ((local && !remote) || (!local && remote)) + status = -EINVAL; + + if (!status && + ((remote->ni_nodenum != local->nd_num) || + (remote->ni_ipv4_port != local->nd_ipv4_port) || + (remote->ni_ipv4_address != local->nd_ipv4_address))) + status = -EINVAL; + + if (status) { + if (remote && !local) + mlog(ML_ERROR, "Domain %s: Node %d (%pI4:%u) " + "registered in joining node %d but not in " + "local node %d\n", qn->qn_domain, + remote->ni_nodenum, + &(remote->ni_ipv4_address), + ntohs(remote->ni_ipv4_port), + qn->qn_nodenum, dlm->node_num); + if (local && !remote) + mlog(ML_ERROR, "Domain %s: Node %d (%pI4:%u) " + "registered in local node %d but not in " + "joining node %d\n", qn->qn_domain, + local->nd_num, &(local->nd_ipv4_address), + ntohs(local->nd_ipv4_port), + dlm->node_num, qn->qn_nodenum); + BUG_ON((!local && !remote)); + } + + if (local) + o2nm_node_put(local); + } + + return status; +} + +static int dlm_send_nodeinfo(struct dlm_ctxt *dlm, unsigned long *node_map) +{ + struct dlm_query_nodeinfo *qn = NULL; + struct o2nm_node *node; + int ret = 0, status, count, i; + + if (find_next_bit(node_map, O2NM_MAX_NODES, 0) >= O2NM_MAX_NODES) + goto bail; + + qn = kzalloc(sizeof(struct dlm_query_nodeinfo), GFP_KERNEL); + if (!qn) { + ret = -ENOMEM; + mlog_errno(ret); + goto bail; + } + + for (i = 0, count = 0; i < O2NM_MAX_NODES; ++i) { + node = o2nm_get_node_by_num(i); + if (!node) + continue; + qn->qn_nodes[count].ni_nodenum = node->nd_num; + qn->qn_nodes[count].ni_ipv4_port = node->nd_ipv4_port; + qn->qn_nodes[count].ni_ipv4_address = node->nd_ipv4_address; + mlog(0, "Node %3d, %pI4:%u\n", node->nd_num, + &(node->nd_ipv4_address), ntohs(node->nd_ipv4_port)); + ++count; + o2nm_node_put(node); + } + + qn->qn_nodenum = dlm->node_num; + qn->qn_numnodes = count; + qn->qn_namelen = strlen(dlm->name); + memcpy(qn->qn_domain, dlm->name, qn->qn_namelen); + + i = -1; + while ((i = find_next_bit(node_map, O2NM_MAX_NODES, + i + 1)) < O2NM_MAX_NODES) { + if (i == dlm->node_num) + continue; + + mlog(0, "Sending nodeinfo to node %d\n", i); + + ret = o2net_send_message(DLM_QUERY_NODEINFO, DLM_MOD_KEY, + qn, sizeof(struct dlm_query_nodeinfo), + i, &status); + if (ret >= 0) + ret = status; + if (ret) { + mlog(ML_ERROR, "node mismatch %d, node %d\n", ret, i); + break; + } + } + +bail: + kfree(qn); + return ret; +} + +static int dlm_query_nodeinfo_handler(struct o2net_msg *msg, u32 len, + void *data, void **ret_data) +{ + struct dlm_query_nodeinfo *qn; + struct dlm_ctxt *dlm = NULL; + int locked = 0, status = -EINVAL; + + qn = (struct dlm_query_nodeinfo *) msg->buf; + + mlog(0, "Node %u queries nodes on domain %s\n", qn->qn_nodenum, + qn->qn_domain); + + spin_lock(&dlm_domain_lock); + dlm = __dlm_lookup_domain_full(qn->qn_domain, qn->qn_namelen); + if (!dlm) { + mlog(ML_ERROR, "Node %d queried nodes on domain %s before " + "join domain\n", qn->qn_nodenum, qn->qn_domain); + goto bail; + } + + spin_lock(&dlm->spinlock); + locked = 1; + if (dlm->joining_node != qn->qn_nodenum) { + mlog(ML_ERROR, "Node %d queried nodes on domain %s but " + "joining node is %d\n", qn->qn_nodenum, qn->qn_domain, + dlm->joining_node); + goto bail; + } + + /* Support for node query was added in 1.1 */ + if (dlm->dlm_locking_proto.pv_major == 1 && + dlm->dlm_locking_proto.pv_minor == 0) { + mlog(ML_ERROR, "Node %d queried nodes on domain %s " + "but active dlm protocol is %d.%d\n", qn->qn_nodenum, + qn->qn_domain, dlm->dlm_locking_proto.pv_major, + dlm->dlm_locking_proto.pv_minor); + goto bail; + } + + status = dlm_match_nodes(dlm, qn); + +bail: + if (locked) + spin_unlock(&dlm->spinlock); + spin_unlock(&dlm_domain_lock); + + return status; +} + static int dlm_cancel_join_handler(struct o2net_msg *msg, u32 len, void *data, void **ret_data) { @@ -1240,6 +1611,20 @@ static int dlm_try_to_join_domain(struct dlm_ctxt *dlm) set_bit(dlm->node_num, dlm->domain_map); spin_unlock(&dlm->spinlock); + /* Support for global heartbeat and node info was added in 1.1 */ + if (dlm_protocol.pv_major > 1 || dlm_protocol.pv_minor > 0) { + status = dlm_send_nodeinfo(dlm, ctxt->yes_resp_map); + if (status) { + mlog_errno(status); + goto bail; + } + status = dlm_send_regions(dlm, ctxt->yes_resp_map); + if (status) { + mlog_errno(status); + goto bail; + } + } + dlm_send_join_asserts(dlm, ctxt->yes_resp_map); /* Joined state *must* be set before the joining node @@ -1806,7 +2191,21 @@ static int dlm_register_net_handlers(void) sizeof(struct dlm_cancel_join), dlm_cancel_join_handler, NULL, NULL, &dlm_join_handlers); + if (status) + goto bail; + + status = o2net_register_handler(DLM_QUERY_REGION, DLM_MOD_KEY, + sizeof(struct dlm_query_region), + dlm_query_region_handler, + NULL, NULL, &dlm_join_handlers); + if (status) + goto bail; + + status = o2net_register_handler(DLM_QUERY_NODEINFO, DLM_MOD_KEY, + sizeof(struct dlm_query_nodeinfo), + dlm_query_nodeinfo_handler, + NULL, NULL, &dlm_join_handlers); bail: if (status < 0) dlm_unregister_net_handlers(); diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c index ffb4c68dafa..f564b0e5f80 100644 --- a/fs/ocfs2/dlm/dlmmaster.c +++ b/fs/ocfs2/dlm/dlmmaster.c @@ -3433,3 +3433,43 @@ void dlm_lockres_release_ast(struct dlm_ctxt *dlm, wake_up(&res->wq); wake_up(&dlm->migration_wq); } + +void dlm_force_free_mles(struct dlm_ctxt *dlm) +{ + int i; + struct hlist_head *bucket; + struct dlm_master_list_entry *mle; + struct hlist_node *tmp, *list; + + /* + * We notified all other nodes that we are exiting the domain and + * marked the dlm state to DLM_CTXT_LEAVING. If any mles are still + * around we force free them and wake any processes that are waiting + * on the mles + */ + spin_lock(&dlm->spinlock); + spin_lock(&dlm->master_lock); + + BUG_ON(dlm->dlm_state != DLM_CTXT_LEAVING); + BUG_ON((find_next_bit(dlm->domain_map, O2NM_MAX_NODES, 0) < O2NM_MAX_NODES)); + + for (i = 0; i < DLM_HASH_BUCKETS; i++) { + bucket = dlm_master_hash(dlm, i); + hlist_for_each_safe(list, tmp, bucket) { + mle = hlist_entry(list, struct dlm_master_list_entry, + master_hash_node); + if (mle->type != DLM_MLE_BLOCK) { + mlog(ML_ERROR, "bad mle: %p\n", mle); + dlm_print_one_mle(mle); + } + atomic_set(&mle->woken, 1); + wake_up(&mle->wq); + + __dlm_unlink_mle(dlm, mle); + __dlm_mle_detach_hb_events(dlm, mle); + __dlm_put_mle(mle); + } + } + spin_unlock(&dlm->master_lock); + spin_unlock(&dlm->spinlock); +} diff --git a/fs/ocfs2/dlmglue.h b/fs/ocfs2/dlmglue.h index d1ce48e1b3d..1d596d8c4a4 100644 --- a/fs/ocfs2/dlmglue.h +++ b/fs/ocfs2/dlmglue.h @@ -84,6 +84,7 @@ enum { OI_LS_PARENT, OI_LS_RENAME1, OI_LS_RENAME2, + OI_LS_REFLINK_TARGET, }; int ocfs2_dlm_init(struct ocfs2_super *osb); diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index 3064feef143..d8408217e3b 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -250,7 +250,7 @@ enum ocfs2_local_alloc_state enum ocfs2_mount_options { - OCFS2_MOUNT_HB_LOCAL = 1 << 0, /* Heartbeat started in local mode */ + OCFS2_MOUNT_HB_LOCAL = 1 << 0, /* Local heartbeat */ OCFS2_MOUNT_BARRIER = 1 << 1, /* Use block barriers */ OCFS2_MOUNT_NOINTR = 1 << 2, /* Don't catch signals */ OCFS2_MOUNT_ERRORS_PANIC = 1 << 3, /* Panic on errors */ @@ -263,9 +263,10 @@ enum ocfs2_mount_options control lists */ OCFS2_MOUNT_USRQUOTA = 1 << 10, /* We support user quotas */ OCFS2_MOUNT_GRPQUOTA = 1 << 11, /* We support group quotas */ - - OCFS2_MOUNT_COHERENCY_BUFFERED = 1 << 12 /* Allow concurrent O_DIRECT - writes */ + OCFS2_MOUNT_COHERENCY_BUFFERED = 1 << 12, /* Allow concurrent O_DIRECT + writes */ + OCFS2_MOUNT_HB_NONE = 1 << 13, /* No heartbeat */ + OCFS2_MOUNT_HB_GLOBAL = 1 << 14, /* Global heartbeat */ }; #define OCFS2_OSB_SOFT_RO 0x0001 @@ -379,6 +380,8 @@ struct ocfs2_super struct ocfs2_alloc_stats alloc_stats; char dev_str[20]; /* "major,minor" of the device */ + u8 osb_stackflags; + char osb_cluster_stack[OCFS2_STACK_LABEL_LEN + 1]; struct ocfs2_cluster_connection *cconn; struct ocfs2_lock_res osb_super_lockres; @@ -612,10 +615,35 @@ static inline int ocfs2_is_soft_readonly(struct ocfs2_super *osb) return ret; } -static inline int ocfs2_userspace_stack(struct ocfs2_super *osb) +static inline int ocfs2_clusterinfo_valid(struct ocfs2_super *osb) { return (osb->s_feature_incompat & - OCFS2_FEATURE_INCOMPAT_USERSPACE_STACK); + (OCFS2_FEATURE_INCOMPAT_USERSPACE_STACK | + OCFS2_FEATURE_INCOMPAT_CLUSTERINFO)); +} + +static inline int ocfs2_userspace_stack(struct ocfs2_super *osb) +{ + if (ocfs2_clusterinfo_valid(osb) && + memcmp(osb->osb_cluster_stack, OCFS2_CLASSIC_CLUSTER_STACK, + OCFS2_STACK_LABEL_LEN)) + return 1; + return 0; +} + +static inline int ocfs2_o2cb_stack(struct ocfs2_super *osb) +{ + if (ocfs2_clusterinfo_valid(osb) && + !memcmp(osb->osb_cluster_stack, OCFS2_CLASSIC_CLUSTER_STACK, + OCFS2_STACK_LABEL_LEN)) + return 1; + return 0; +} + +static inline int ocfs2_cluster_o2cb_global_heartbeat(struct ocfs2_super *osb) +{ + return ocfs2_o2cb_stack(osb) && + (osb->osb_stackflags & OCFS2_CLUSTER_O2CB_GLOBAL_HEARTBEAT); } static inline int ocfs2_mount_local(struct ocfs2_super *osb) diff --git a/fs/ocfs2/ocfs2_fs.h b/fs/ocfs2/ocfs2_fs.h index 723b20dac41..c2e4f8222e2 100644 --- a/fs/ocfs2/ocfs2_fs.h +++ b/fs/ocfs2/ocfs2_fs.h @@ -101,7 +101,8 @@ | OCFS2_FEATURE_INCOMPAT_META_ECC \ | OCFS2_FEATURE_INCOMPAT_INDEXED_DIRS \ | OCFS2_FEATURE_INCOMPAT_REFCOUNT_TREE \ - | OCFS2_FEATURE_INCOMPAT_DISCONTIG_BG) + | OCFS2_FEATURE_INCOMPAT_DISCONTIG_BG \ + | OCFS2_FEATURE_INCOMPAT_CLUSTERINFO) #define OCFS2_FEATURE_RO_COMPAT_SUPP (OCFS2_FEATURE_RO_COMPAT_UNWRITTEN \ | OCFS2_FEATURE_RO_COMPAT_USRQUOTA \ | OCFS2_FEATURE_RO_COMPAT_GRPQUOTA) @@ -170,6 +171,13 @@ #define OCFS2_FEATURE_INCOMPAT_DISCONTIG_BG 0x2000 /* + * Incompat bit to indicate useable clusterinfo with stackflags for all + * cluster stacks (userspace adnd o2cb). If this bit is set, + * INCOMPAT_USERSPACE_STACK becomes superfluous and thus should not be set. + */ +#define OCFS2_FEATURE_INCOMPAT_CLUSTERINFO 0x4000 + +/* * backup superblock flag is used to indicate that this volume * has backup superblocks. */ @@ -235,18 +243,31 @@ #define OCFS2_HAS_REFCOUNT_FL (0x0010) /* Inode attributes, keep in sync with EXT2 */ -#define OCFS2_SECRM_FL (0x00000001) /* Secure deletion */ -#define OCFS2_UNRM_FL (0x00000002) /* Undelete */ -#define OCFS2_COMPR_FL (0x00000004) /* Compress file */ -#define OCFS2_SYNC_FL (0x00000008) /* Synchronous updates */ -#define OCFS2_IMMUTABLE_FL (0x00000010) /* Immutable file */ -#define OCFS2_APPEND_FL (0x00000020) /* writes to file may only append */ -#define OCFS2_NODUMP_FL (0x00000040) /* do not dump file */ -#define OCFS2_NOATIME_FL (0x00000080) /* do not update atime */ -#define OCFS2_DIRSYNC_FL (0x00010000) /* dirsync behaviour (directories only) */ - -#define OCFS2_FL_VISIBLE (0x000100FF) /* User visible flags */ -#define OCFS2_FL_MODIFIABLE (0x000100FF) /* User modifiable flags */ +#define OCFS2_SECRM_FL FS_SECRM_FL /* Secure deletion */ +#define OCFS2_UNRM_FL FS_UNRM_FL /* Undelete */ +#define OCFS2_COMPR_FL FS_COMPR_FL /* Compress file */ +#define OCFS2_SYNC_FL FS_SYNC_FL /* Synchronous updates */ +#define OCFS2_IMMUTABLE_FL FS_IMMUTABLE_FL /* Immutable file */ +#define OCFS2_APPEND_FL FS_APPEND_FL /* writes to file may only append */ +#define OCFS2_NODUMP_FL FS_NODUMP_FL /* do not dump file */ +#define OCFS2_NOATIME_FL FS_NOATIME_FL /* do not update atime */ +/* Reserved for compression usage... */ +#define OCFS2_DIRTY_FL FS_DIRTY_FL +#define OCFS2_COMPRBLK_FL FS_COMPRBLK_FL /* One or more compressed clusters */ +#define OCFS2_NOCOMP_FL FS_NOCOMP_FL /* Don't compress */ +#define OCFS2_ECOMPR_FL FS_ECOMPR_FL /* Compression error */ +/* End compression flags --- maybe not all used */ +#define OCFS2_BTREE_FL FS_BTREE_FL /* btree format dir */ +#define OCFS2_INDEX_FL FS_INDEX_FL /* hash-indexed directory */ +#define OCFS2_IMAGIC_FL FS_IMAGIC_FL /* AFS directory */ +#define OCFS2_JOURNAL_DATA_FL FS_JOURNAL_DATA_FL /* Reserved for ext3 */ +#define OCFS2_NOTAIL_FL FS_NOTAIL_FL /* file tail should not be merged */ +#define OCFS2_DIRSYNC_FL FS_DIRSYNC_FL /* dirsync behaviour (directories only) */ +#define OCFS2_TOPDIR_FL FS_TOPDIR_FL /* Top of directory hierarchies*/ +#define OCFS2_RESERVED_FL FS_RESERVED_FL /* reserved for ext2 lib */ + +#define OCFS2_FL_VISIBLE FS_FL_USER_VISIBLE /* User visible flags */ +#define OCFS2_FL_MODIFIABLE FS_FL_USER_MODIFIABLE /* User modifiable flags */ /* * Extent record flags (e_node.leaf.flags) @@ -279,10 +300,13 @@ #define OCFS2_VOL_UUID_LEN 16 #define OCFS2_MAX_VOL_LABEL_LEN 64 -/* The alternate, userspace stack fields */ +/* The cluster stack fields */ #define OCFS2_STACK_LABEL_LEN 4 #define OCFS2_CLUSTER_NAME_LEN 16 +/* Classic (historically speaking) cluster stack */ +#define OCFS2_CLASSIC_CLUSTER_STACK "o2cb" + /* Journal limits (in bytes) */ #define OCFS2_MIN_JOURNAL_SIZE (4 * 1024 * 1024) @@ -292,6 +316,11 @@ */ #define OCFS2_MIN_XATTR_INLINE_SIZE 256 +/* + * Cluster info flags (ocfs2_cluster_info.ci_stackflags) + */ +#define OCFS2_CLUSTER_O2CB_GLOBAL_HEARTBEAT (0x01) + struct ocfs2_system_inode_info { char *si_name; int si_iflags; @@ -352,6 +381,7 @@ static struct ocfs2_system_inode_info ocfs2_system_inodes[NUM_SYSTEM_INODES] = { /* Parameter passed from mount.ocfs2 to module */ #define OCFS2_HB_NONE "heartbeat=none" #define OCFS2_HB_LOCAL "heartbeat=local" +#define OCFS2_HB_GLOBAL "heartbeat=global" /* * OCFS2 directory file types. Only the low 3 bits are used. The @@ -558,9 +588,21 @@ struct ocfs2_slot_map_extended { */ }; +/* + * ci_stackflags is only valid if the incompat bit + * OCFS2_FEATURE_INCOMPAT_CLUSTERINFO is set. + */ struct ocfs2_cluster_info { /*00*/ __u8 ci_stack[OCFS2_STACK_LABEL_LEN]; - __le32 ci_reserved; + union { + __le32 ci_reserved; + struct { + __u8 ci_stackflags; + __u8 ci_reserved1; + __u8 ci_reserved2; + __u8 ci_reserved3; + }; + }; /*08*/ __u8 ci_cluster[OCFS2_CLUSTER_NAME_LEN]; /*18*/ }; @@ -597,9 +639,9 @@ struct ocfs2_super_block { * group header */ /*50*/ __u8 s_label[OCFS2_MAX_VOL_LABEL_LEN]; /* Label for mounting, etc. */ /*90*/ __u8 s_uuid[OCFS2_VOL_UUID_LEN]; /* 128-bit uuid */ -/*A0*/ struct ocfs2_cluster_info s_cluster_info; /* Selected userspace - stack. Only valid - with INCOMPAT flag. */ +/*A0*/ struct ocfs2_cluster_info s_cluster_info; /* Only valid if either + userspace or clusterinfo + INCOMPAT flag set. */ /*B8*/ __le16 s_xattr_inline_size; /* extended attribute inline size for this fs*/ __le16 s_reserved0; diff --git a/fs/ocfs2/ocfs2_ioctl.h b/fs/ocfs2/ocfs2_ioctl.h index 9bc53549986..b46f39bf743 100644 --- a/fs/ocfs2/ocfs2_ioctl.h +++ b/fs/ocfs2/ocfs2_ioctl.h @@ -23,10 +23,10 @@ /* * ioctl commands */ -#define OCFS2_IOC_GETFLAGS _IOR('f', 1, long) -#define OCFS2_IOC_SETFLAGS _IOW('f', 2, long) -#define OCFS2_IOC32_GETFLAGS _IOR('f', 1, int) -#define OCFS2_IOC32_SETFLAGS _IOW('f', 2, int) +#define OCFS2_IOC_GETFLAGS FS_IOC_GETFLAGS +#define OCFS2_IOC_SETFLAGS FS_IOC_SETFLAGS +#define OCFS2_IOC32_GETFLAGS FS_IOC32_GETFLAGS +#define OCFS2_IOC32_SETFLAGS FS_IOC32_SETFLAGS /* * Space reservation / allocation / free ioctls and argument structure diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c index a120cfcf69b..b5f9160e93e 100644 --- a/fs/ocfs2/refcounttree.c +++ b/fs/ocfs2/refcounttree.c @@ -4240,8 +4240,9 @@ static int __ocfs2_reflink(struct dentry *old_dentry, goto out; } - mutex_lock(&new_inode->i_mutex); - ret = ocfs2_inode_lock(new_inode, &new_bh, 1); + mutex_lock_nested(&new_inode->i_mutex, I_MUTEX_CHILD); + ret = ocfs2_inode_lock_nested(new_inode, &new_bh, 1, + OI_LS_REFLINK_TARGET); if (ret) { mlog_errno(ret); goto out_unlock; diff --git a/fs/ocfs2/reservations.c b/fs/ocfs2/reservations.c index d8b6e4259b8..3e78db361bc 100644 --- a/fs/ocfs2/reservations.c +++ b/fs/ocfs2/reservations.c @@ -732,25 +732,23 @@ int ocfs2_resmap_resv_bits(struct ocfs2_reservation_map *resmap, struct ocfs2_alloc_reservation *resv, int *cstart, int *clen) { - unsigned int wanted = *clen; - if (resv == NULL || ocfs2_resmap_disabled(resmap)) return -ENOSPC; spin_lock(&resv_lock); - /* - * We don't want to over-allocate for temporary - * windows. Otherwise, we run the risk of fragmenting the - * allocation space. - */ - wanted = ocfs2_resv_window_bits(resmap, resv); - if ((resv->r_flags & OCFS2_RESV_FLAG_TMP) || wanted < *clen) - wanted = *clen; - if (ocfs2_resv_empty(resv)) { - mlog(0, "empty reservation, find new window\n"); + /* + * We don't want to over-allocate for temporary + * windows. Otherwise, we run the risk of fragmenting the + * allocation space. + */ + unsigned int wanted = ocfs2_resv_window_bits(resmap, resv); + if ((resv->r_flags & OCFS2_RESV_FLAG_TMP) || wanted < *clen) + wanted = *clen; + + mlog(0, "empty reservation, find new window\n"); /* * Try to get a window here. If it works, we must fall * through and test the bitmap . This avoids some diff --git a/fs/ocfs2/stack_o2cb.c b/fs/ocfs2/stack_o2cb.c index 0d3049f696c..19965b00c43 100644 --- a/fs/ocfs2/stack_o2cb.c +++ b/fs/ocfs2/stack_o2cb.c @@ -283,6 +283,8 @@ static int o2cb_cluster_connect(struct ocfs2_cluster_connection *conn) /* for now we only have one cluster/node, make sure we see it * in the heartbeat universe */ if (!o2hb_check_local_node_heartbeating()) { + if (o2hb_global_heartbeat_active()) + mlog(ML_ERROR, "Global heartbeat not started\n"); rc = -EINVAL; goto out; } diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c index 64f2c50a1c3..5fed60de763 100644 --- a/fs/ocfs2/suballoc.c +++ b/fs/ocfs2/suballoc.c @@ -357,7 +357,7 @@ out: static void ocfs2_bg_discontig_add_extent(struct ocfs2_super *osb, struct ocfs2_group_desc *bg, struct ocfs2_chain_list *cl, - u64 p_blkno, u32 clusters) + u64 p_blkno, unsigned int clusters) { struct ocfs2_extent_list *el = &bg->bg_list; struct ocfs2_extent_rec *rec; @@ -369,7 +369,7 @@ static void ocfs2_bg_discontig_add_extent(struct ocfs2_super *osb, rec->e_blkno = cpu_to_le64(p_blkno); rec->e_cpos = cpu_to_le32(le16_to_cpu(bg->bg_bits) / le16_to_cpu(cl->cl_bpc)); - rec->e_leaf_clusters = cpu_to_le32(clusters); + rec->e_leaf_clusters = cpu_to_le16(clusters); le16_add_cpu(&bg->bg_bits, clusters * le16_to_cpu(cl->cl_bpc)); le16_add_cpu(&bg->bg_free_bits_count, clusters * le16_to_cpu(cl->cl_bpc)); diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 9122d59f812..a8a0ca44f88 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -162,6 +162,7 @@ enum { Opt_nointr, Opt_hb_none, Opt_hb_local, + Opt_hb_global, Opt_data_ordered, Opt_data_writeback, Opt_atime_quantum, @@ -192,6 +193,7 @@ static const match_table_t tokens = { {Opt_nointr, "nointr"}, {Opt_hb_none, OCFS2_HB_NONE}, {Opt_hb_local, OCFS2_HB_LOCAL}, + {Opt_hb_global, OCFS2_HB_GLOBAL}, {Opt_data_ordered, "data=ordered"}, {Opt_data_writeback, "data=writeback"}, {Opt_atime_quantum, "atime_quantum=%u"}, @@ -626,6 +628,7 @@ static int ocfs2_remount(struct super_block *sb, int *flags, char *data) int ret = 0; struct mount_options parsed_options; struct ocfs2_super *osb = OCFS2_SB(sb); + u32 tmp; lock_kernel(); @@ -635,8 +638,9 @@ static int ocfs2_remount(struct super_block *sb, int *flags, char *data) goto out; } - if ((osb->s_mount_opt & OCFS2_MOUNT_HB_LOCAL) != - (parsed_options.mount_opt & OCFS2_MOUNT_HB_LOCAL)) { + tmp = OCFS2_MOUNT_HB_LOCAL | OCFS2_MOUNT_HB_GLOBAL | + OCFS2_MOUNT_HB_NONE; + if ((osb->s_mount_opt & tmp) != (parsed_options.mount_opt & tmp)) { ret = -EINVAL; mlog(ML_ERROR, "Cannot change heartbeat mode on remount\n"); goto out; @@ -827,23 +831,29 @@ bail: static int ocfs2_verify_heartbeat(struct ocfs2_super *osb) { - if (ocfs2_mount_local(osb)) { - if (osb->s_mount_opt & OCFS2_MOUNT_HB_LOCAL) { + u32 hb_enabled = OCFS2_MOUNT_HB_LOCAL | OCFS2_MOUNT_HB_GLOBAL; + + if (osb->s_mount_opt & hb_enabled) { + if (ocfs2_mount_local(osb)) { mlog(ML_ERROR, "Cannot heartbeat on a locally " "mounted device.\n"); return -EINVAL; } - } - - if (ocfs2_userspace_stack(osb)) { - if (osb->s_mount_opt & OCFS2_MOUNT_HB_LOCAL) { + if (ocfs2_userspace_stack(osb)) { mlog(ML_ERROR, "Userspace stack expected, but " "o2cb heartbeat arguments passed to mount\n"); return -EINVAL; } + if (((osb->s_mount_opt & OCFS2_MOUNT_HB_GLOBAL) && + !ocfs2_cluster_o2cb_global_heartbeat(osb)) || + ((osb->s_mount_opt & OCFS2_MOUNT_HB_LOCAL) && + ocfs2_cluster_o2cb_global_heartbeat(osb))) { + mlog(ML_ERROR, "Mismatching o2cb heartbeat modes\n"); + return -EINVAL; + } } - if (!(osb->s_mount_opt & OCFS2_MOUNT_HB_LOCAL)) { + if (!(osb->s_mount_opt & hb_enabled)) { if (!ocfs2_mount_local(osb) && !ocfs2_is_hard_readonly(osb) && !ocfs2_userspace_stack(osb)) { mlog(ML_ERROR, "Heartbeat has to be started to mount " @@ -1309,6 +1319,7 @@ static int ocfs2_parse_options(struct super_block *sb, { int status; char *p; + u32 tmp; mlog_entry("remount: %d, options: \"%s\"\n", is_remount, options ? options : "(none)"); @@ -1340,7 +1351,10 @@ static int ocfs2_parse_options(struct super_block *sb, mopt->mount_opt |= OCFS2_MOUNT_HB_LOCAL; break; case Opt_hb_none: - mopt->mount_opt &= ~OCFS2_MOUNT_HB_LOCAL; + mopt->mount_opt |= OCFS2_MOUNT_HB_NONE; + break; + case Opt_hb_global: + mopt->mount_opt |= OCFS2_MOUNT_HB_GLOBAL; break; case Opt_barrier: if (match_int(&args[0], &option)) { @@ -1501,6 +1515,15 @@ static int ocfs2_parse_options(struct super_block *sb, } } + /* Ensure only one heartbeat mode */ + tmp = mopt->mount_opt & (OCFS2_MOUNT_HB_LOCAL | OCFS2_MOUNT_HB_GLOBAL | + OCFS2_MOUNT_HB_NONE); + if (hweight32(tmp) != 1) { + mlog(ML_ERROR, "Invalid heartbeat mount options\n"); + status = 0; + goto bail; + } + status = 1; bail: @@ -1514,10 +1537,14 @@ static int ocfs2_show_options(struct seq_file *s, struct vfsmount *mnt) unsigned long opts = osb->s_mount_opt; unsigned int local_alloc_megs; - if (opts & OCFS2_MOUNT_HB_LOCAL) - seq_printf(s, ",_netdev,heartbeat=local"); - else - seq_printf(s, ",heartbeat=none"); + if (opts & (OCFS2_MOUNT_HB_LOCAL | OCFS2_MOUNT_HB_GLOBAL)) { + seq_printf(s, ",_netdev"); + if (opts & OCFS2_MOUNT_HB_LOCAL) + seq_printf(s, ",%s", OCFS2_HB_LOCAL); + else + seq_printf(s, ",%s", OCFS2_HB_GLOBAL); + } else + seq_printf(s, ",%s", OCFS2_HB_NONE); if (opts & OCFS2_MOUNT_NOINTR) seq_printf(s, ",nointr"); @@ -2209,7 +2236,9 @@ static int ocfs2_initialize_super(struct super_block *sb, goto bail; } - if (ocfs2_userspace_stack(osb)) { + if (ocfs2_clusterinfo_valid(osb)) { + osb->osb_stackflags = + OCFS2_RAW_SB(di)->s_cluster_info.ci_stackflags; memcpy(osb->osb_cluster_stack, OCFS2_RAW_SB(di)->s_cluster_info.ci_stack, OCFS2_STACK_LABEL_LEN); diff --git a/fs/ocfs2/symlink.c b/fs/ocfs2/symlink.c index 32499d213fc..9975457c981 100644 --- a/fs/ocfs2/symlink.c +++ b/fs/ocfs2/symlink.c @@ -128,7 +128,7 @@ static void *ocfs2_fast_follow_link(struct dentry *dentry, } /* Fast symlinks can't be large */ - len = strlen(target); + len = strnlen(target, ocfs2_fast_symlink_chars(inode->i_sb)); link = kzalloc(len + 1, GFP_NOFS); if (!link) { status = -ENOMEM; diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index d03469f6180..06fa5e77c40 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -1286,13 +1286,11 @@ int ocfs2_xattr_get_nolock(struct inode *inode, xis.inode_bh = xbs.inode_bh = di_bh; di = (struct ocfs2_dinode *)di_bh->b_data; - down_read(&oi->ip_xattr_sem); ret = ocfs2_xattr_ibody_get(inode, name_index, name, buffer, buffer_size, &xis); if (ret == -ENODATA && di->i_xattr_loc) ret = ocfs2_xattr_block_get(inode, name_index, name, buffer, buffer_size, &xbs); - up_read(&oi->ip_xattr_sem); return ret; } @@ -1316,8 +1314,10 @@ static int ocfs2_xattr_get(struct inode *inode, mlog_errno(ret); return ret; } + down_read(&OCFS2_I(inode)->ip_xattr_sem); ret = ocfs2_xattr_get_nolock(inode, di_bh, name_index, name, buffer, buffer_size); + up_read(&OCFS2_I(inode)->ip_xattr_sem); ocfs2_inode_unlock(inode, 0); diff --git a/fs/proc/base.c b/fs/proc/base.c index a1c43e7c8a7..8e4addaa542 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -2675,7 +2675,7 @@ static const struct pid_entry tgid_base_stuff[] = { INF("auxv", S_IRUSR, proc_pid_auxv), ONE("status", S_IRUGO, proc_pid_status), ONE("personality", S_IRUSR, proc_pid_personality), - INF("limits", S_IRUSR, proc_pid_limits), + INF("limits", S_IRUGO, proc_pid_limits), #ifdef CONFIG_SCHED_DEBUG REG("sched", S_IRUGO|S_IWUSR, proc_pid_sched_operations), #endif @@ -3011,7 +3011,7 @@ static const struct pid_entry tid_base_stuff[] = { INF("auxv", S_IRUSR, proc_pid_auxv), ONE("status", S_IRUGO, proc_pid_status), ONE("personality", S_IRUSR, proc_pid_personality), - INF("limits", S_IRUSR, proc_pid_limits), + INF("limits", S_IRUGO, proc_pid_limits), #ifdef CONFIG_SCHED_DEBUG REG("sched", S_IRUGO|S_IWUSR, proc_pid_sched_operations), #endif diff --git a/fs/proc/page.c b/fs/proc/page.c index 180cf5a0bd6..3b8b4566033 100644 --- a/fs/proc/page.c +++ b/fs/proc/page.c @@ -146,7 +146,7 @@ u64 stable_page_flags(struct page *page) u |= kpf_copy_bit(k, KPF_HWPOISON, PG_hwpoison); #endif -#ifdef CONFIG_IA64_UNCACHED_ALLOCATOR +#ifdef CONFIG_ARCH_USES_PG_UNCACHED u |= kpf_copy_bit(k, KPF_UNCACHED, PG_uncached); #endif diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 271afc48b9a..1dbca4e8cc1 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -363,13 +363,13 @@ static int smaps_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end, mss->referenced += PAGE_SIZE; mapcount = page_mapcount(page); if (mapcount >= 2) { - if (pte_dirty(ptent)) + if (pte_dirty(ptent) || PageDirty(page)) mss->shared_dirty += PAGE_SIZE; else mss->shared_clean += PAGE_SIZE; mss->pss += (PAGE_SIZE << PSS_SHIFT) / mapcount; } else { - if (pte_dirty(ptent)) + if (pte_dirty(ptent) || PageDirty(page)) mss->private_dirty += PAGE_SIZE; else mss->private_clean += PAGE_SIZE; diff --git a/fs/proc/vmcore.c b/fs/proc/vmcore.c index 91c817ff02c..2367fb3f70b 100644 --- a/fs/proc/vmcore.c +++ b/fs/proc/vmcore.c @@ -163,7 +163,7 @@ static ssize_t read_vmcore(struct file *file, char __user *buffer, static const struct file_operations proc_vmcore_operations = { .read = read_vmcore, - .llseek = generic_file_llseek, + .llseek = default_llseek, }; static struct vmcore* __init get_new_element(void) diff --git a/fs/reiserfs/ioctl.c b/fs/reiserfs/ioctl.c index f53505de071..5cbb81e134a 100644 --- a/fs/reiserfs/ioctl.c +++ b/fs/reiserfs/ioctl.c @@ -170,6 +170,7 @@ int reiserfs_prepare_write(struct file *f, struct page *page, int reiserfs_unpack(struct inode *inode, struct file *filp) { int retval = 0; + int depth; int index; struct page *page; struct address_space *mapping; @@ -188,8 +189,8 @@ int reiserfs_unpack(struct inode *inode, struct file *filp) /* we need to make sure nobody is changing the file size beneath ** us */ - mutex_lock(&inode->i_mutex); - reiserfs_write_lock(inode->i_sb); + reiserfs_mutex_lock_safe(&inode->i_mutex, inode->i_sb); + depth = reiserfs_write_lock_once(inode->i_sb); write_from = inode->i_size & (blocksize - 1); /* if we are on a block boundary, we are already unpacked. */ @@ -224,6 +225,6 @@ int reiserfs_unpack(struct inode *inode, struct file *filp) out: mutex_unlock(&inode->i_mutex); - reiserfs_write_unlock(inode->i_sb); + reiserfs_write_unlock_once(inode->i_sb, depth); return retval; } diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index d72cf2bb054..286e36e21da 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -1932,7 +1932,8 @@ xfs_buf_init(void) if (!xfs_buf_zone) goto out; - xfslogd_workqueue = create_workqueue("xfslogd"); + xfslogd_workqueue = alloc_workqueue("xfslogd", + WQ_RESCUER | WQ_HIGHPRI, 1); if (!xfslogd_workqueue) goto out_free_buf_zone; diff --git a/fs/xfs/linux-2.6/xfs_ioctl.c b/fs/xfs/linux-2.6/xfs_ioctl.c index 4fec427b83e..3b9e626f7cd 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl.c +++ b/fs/xfs/linux-2.6/xfs_ioctl.c @@ -785,6 +785,8 @@ xfs_ioc_fsgetxattr( { struct fsxattr fa; + memset(&fa, 0, sizeof(struct fsxattr)); + xfs_ilock(ip, XFS_ILOCK_SHARED); fa.fsx_xflags = xfs_ip2xflags(ip); fa.fsx_extsize = ip->i_d.di_extsize << ip->i_mount->m_sb.sb_blocklog; diff --git a/fs/xfs/xfs_log_cil.c b/fs/xfs/xfs_log_cil.c index ed575fb4b49..7e206fc1fa3 100644 --- a/fs/xfs/xfs_log_cil.c +++ b/fs/xfs/xfs_log_cil.c @@ -405,9 +405,15 @@ xlog_cil_push( new_ctx = kmem_zalloc(sizeof(*new_ctx), KM_SLEEP|KM_NOFS); new_ctx->ticket = xlog_cil_ticket_alloc(log); - /* lock out transaction commit, but don't block on background push */ + /* + * Lock out transaction commit, but don't block for background pushes + * unless we are well over the CIL space limit. See the definition of + * XLOG_CIL_HARD_SPACE_LIMIT() for the full explanation of the logic + * used here. + */ if (!down_write_trylock(&cil->xc_ctx_lock)) { - if (!push_seq) + if (!push_seq && + cil->xc_ctx->space_used < XLOG_CIL_HARD_SPACE_LIMIT(log)) goto out_free_ticket; down_write(&cil->xc_ctx_lock); } @@ -422,7 +428,7 @@ xlog_cil_push( goto out_skip; /* check for a previously pushed seqeunce */ - if (push_seq < cil->xc_ctx->sequence) + if (push_seq && push_seq < cil->xc_ctx->sequence) goto out_skip; /* diff --git a/fs/xfs/xfs_log_priv.h b/fs/xfs/xfs_log_priv.h index ced52b98b32..edcdfe01617 100644 --- a/fs/xfs/xfs_log_priv.h +++ b/fs/xfs/xfs_log_priv.h @@ -426,13 +426,13 @@ struct xfs_cil { }; /* - * The amount of log space we should the CIL to aggregate is difficult to size. - * Whatever we chose we have to make we can get a reservation for the log space - * effectively, that it is large enough to capture sufficient relogging to - * reduce log buffer IO significantly, but it is not too large for the log or - * induces too much latency when writing out through the iclogs. We track both - * space consumed and the number of vectors in the checkpoint context, so we - * need to decide which to use for limiting. + * The amount of log space we allow the CIL to aggregate is difficult to size. + * Whatever we choose, we have to make sure we can get a reservation for the + * log space effectively, that it is large enough to capture sufficient + * relogging to reduce log buffer IO significantly, but it is not too large for + * the log or induces too much latency when writing out through the iclogs. We + * track both space consumed and the number of vectors in the checkpoint + * context, so we need to decide which to use for limiting. * * Every log buffer we write out during a push needs a header reserved, which * is at least one sector and more for v2 logs. Hence we need a reservation of @@ -459,16 +459,21 @@ struct xfs_cil { * checkpoint transaction ticket is specific to the checkpoint context, rather * than the CIL itself. * - * With dynamic reservations, we can basically make up arbitrary limits for the - * checkpoint size so long as they don't violate any other size rules. Hence - * the initial maximum size for the checkpoint transaction will be set to a - * quarter of the log or 8MB, which ever is smaller. 8MB is an arbitrary limit - * right now based on the latency of writing out a large amount of data through - * the circular iclog buffers. + * With dynamic reservations, we can effectively make up arbitrary limits for + * the checkpoint size so long as they don't violate any other size rules. + * Recovery imposes a rule that no transaction exceed half the log, so we are + * limited by that. Furthermore, the log transaction reservation subsystem + * tries to keep 25% of the log free, so we need to keep below that limit or we + * risk running out of free log space to start any new transactions. + * + * In order to keep background CIL push efficient, we will set a lower + * threshold at which background pushing is attempted without blocking current + * transaction commits. A separate, higher bound defines when CIL pushes are + * enforced to ensure we stay within our maximum checkpoint size bounds. + * threshold, yet give us plenty of space for aggregation on large logs. */ - -#define XLOG_CIL_SPACE_LIMIT(log) \ - (min((log->l_logsize >> 2), (8 * 1024 * 1024))) +#define XLOG_CIL_SPACE_LIMIT(log) (log->l_logsize >> 3) +#define XLOG_CIL_HARD_SPACE_LIMIT(log) (3 * (log->l_logsize >> 4)) /* * The reservation head lsn is not made up of a cycle number and block number. |